UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Hydroformylation of glycals 1964

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
UBC_1965_A1 A3.pdf
UBC_1965_A1 A3.pdf [ 10.74MB ]
Metadata
JSON: 1.0062109.json
JSON-LD: 1.0062109+ld.json
RDF/XML (Pretty): 1.0062109.xml
RDF/JSON: 1.0062109+rdf.json
Turtle: 1.0062109+rdf-turtle.txt
N-Triples: 1.0062109+rdf-ntriples.txt
Citation
1.0062109.ris

Full Text

HYDROFORMYLATION OP GLYCALS by DEREK ABSON B.Sc, University of Birmingham, 1952 M.Sc., University of Birmingham, 1961 •A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY i n the Department of Chemistry We accept t h i s thesis as conforming to the required standard September, 1964 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the U n i v e r s i t y of • B r i t i s h Columbia, I agree that the L i b r a r y s h a l l make' i t f r e e l y a v a i l a b l e f o r reference and study. I f u r t h e r agree that per- m i s s i o n f o r extensive copying of t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the Head of my Department or by h i s r e p r e s e n t a t i v e s . I t i s understood t h a t . c o p y i n g or p u b l i - c a t i o n of t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be allowed, without-my w r i t t e n p e r mission. Department of C k £Hyu sJVty The U n i v e r s i t y of B r i t i s h Columbia, Vancouver 8, Canada Date 2. N W Y / I \<U>Sr ACKNOWLEDGEMENTS I wish to express my gratitude to Dr. A. Rosenthal f o r his constant interest and encouragement during the course of t h i s research. The assistance of Dr. L. D. Ha l l i n the measurement and interpretation of n.m.r. spectra i s g r a t e f u l l y acknowledged. Thanks are also due to Drs. E. von Rudloff and P. A. J . Gorin of the P r a i r i e Regional Laboratory, N.R.C., Saskatoon, f o r providing reference samples of cer t a i n compounds. i l l ABSTRACT The reaction of 3>4-di-0-acetyl>i|D-xylal with 3 moles of synthesis gas (CO + 2Hg) under oxo conditions gave pre- dominantly two isomeric 2 , 3-di-0-acetyl-l , 5-anhydro - 4-deoxy h e x i t o l s , by addition of a hydroxymethyl group at C-l of the g l y c a l . The structures of the two polyols, obtained by deaeetylation of the reaction product and f r a c t i o n a t i o n by paper p a r t i t i o n chromatography, were completely esta- b l i s h e d . Formation of a pair of enantiomeric t r i o l ethers by periodate cleavage and sodium borohydride reduction of each polyol showed that they were l ,5-anhydro -4-deoxy- h e x i t o l s , having unbranched carbon skeletons, t h i s also being shown by the proton resonance positions and inten- s i t i e s i n the n.m.r. spectra of the polyols. One of the enantiomeric t r i o l ethers, having the L-configuration, was prepared from a carbohydrate of known structure, 1,4-anhydro-5-deoxy-B-arabino-hex i t o1,.thereby e s t a b l i s h - ing the configurations at C-5 of the two isomeric 1 , 5 - anhydro -4-deoxy-hexitols. Assignments of the D-arabino- and L-xylo- configurations to the two isomers c o n f l i c t e d " ~~ ftp with re s u l t s of Qorin , who had previously assigned the D-arablno- configurations to a 1,5-anhydro-4-deoxy-hexi161 which did not resemble either of our compounds. That these were the D-arabino- and,L-xylo- isomers of 1 ,5-anhydro -4- i v deoxy-hexitol was proved by t h e i r conversion into a pair of isomeric l , 5-anhydro - 4 , 6-dideoxy-hexitols which were i d e n t i c a l with those obtained by the reaction of 3 , 4-di-O- acetyl - 2-deoxy-D-xylopyranosyl chloride with methyl mag- nesium bromide, both series of reactions allowing no pos- s i b i l i t y of configurational inversions. The polyol de- scribed by Gorin was subsequently shown to be the alte r n a - t i v e trans isomer, 1 ,5-anhydro -4-deoxy-D-xylo-hexitol. A concurrent study of the structures of the two anhydrodeoxyhexitols was made by nuclear magnetic resonance, and the stereochemistry of the L-xylo- isomer could be assigned from the m u l t i p l i c i t i e s of the C-4 proton signals. The single C-4 proton i n the deuterated analogue of the L-xylo- isomer (prepared by reacting 3 , 4-di-O-acetyl-D- x y l a l with carbon monoxide and deuterium) was shown to be equatorial by i t s resonance p o s i t i o n , and i t s m u l t i p l i c i t y on deuterium-hydrogen decoupling, t h i s providing evidence for c i s - addition to the double bond of the gl y e a l on hydroformylation. The oxo reaction of 3 , 4 , 6 - t r i - O - a c e t y l - D - g a l a c t a l has been reinvestigated, and found to be e n t i r e l y analogous to those of other g l y c a l s , giving, on deacetylation, a mixture of 2,6-anhydro-3-deoxy-D-galacto- and D-talo- h e p t i t o l s . These were i s o l a t e d and characterised, and t h e i r V stereochemistry established by c o r r e l a t i o n with the D-gluco- isomer, whose structure has been proved by X-ray analysis. The reaction of 3 , 4-di-O-acetyl-D-xylal under hydro- formylation conditions, leading to the formation of alde- hydes rather than alcohols, has been investigated. From the reaction of the g l y c a l with 2 moles of synthesis gas, two isomeric 4 , 5-di - 0-acety1 - 2 , 6-anhydro - 3-deoxy-aldehydo- hexoses were is o l a t e d as t h e i r c r y s t a l l i n e 2 , 4 - d i n i t r o - phenylhydrazones. These were i d e n t i f i e d by conversion of one of them, 4 , 5-dl - 0-acety1 - 2 , 6-anhydro - 3-deoxy-aldehydo- D-lyxo-hexose, to 1 ,5-anhydro-4-deoxy-D-arabino-hexitol, whose structure had been established previously. The two aldehydo-hexoses were also obtained when a mixture of 2 , 3 - di-Q-acetyl -1 ,5-anhydro -4-deoxy-D-arablno- and L-xylo- hexitols were reacted with dimethylsulphoxide and N,N'- dicyclohexylcarbodiimide. v i CONTENTS Page GENERAL INTRODUCTION 1 The Oxo Reaction 1 ( i ) Nature of the Catalyst 4 ( l i ) Mechanism of the Oxo Reaction 5 ( i l l ) E f f e c t of O l e f i n Structure on Hydro- formylation 10 The Glycals 16 (i) Structure of the Glycals ^ 16 ( i i ) Preparation of the Glycals 20 ( i l l ) Reactions of the Glycals 23 The Oxo Reaction of Glycals 30 DISCUSSION 32 A. Hydroxymethylatioh of 3 , 4-di-O-acetyl-D-xylal 32 (i) Reactants and Reaction Conditions 33 ( i i ) Fractionation and Characterisation of Reaction Products 37 ( l i i ) I d e n t i f i c a t i o n of Fractions I and II 4 l (iv) Configurations of Fractions I and II 52 (v) I d e n t i t i e s of Polyols X and Y 77 (vl) Proton Magnetic Resonance and Stereo- chemistry of Fractions I and II 78 v i i Page B. Anhydrodeoxyheptitols from 3 , 4 , 6 v t r i - 0 - a e e t y l - D-galactal 89 (i) Reaction Conditions QO ( i i ) Characterisation of Fractions A and B 93 ( i i i ) Structures of Fractions A and B 94 (iv) Stereochemistry of Fractions A and B 99 (v) Steric Aspects of Hydroxymethylation 117 C. Hydroformylation of 3 , 4-di-O-acetyl-D-xylal 120 (i) Reaction Conditions and Product I s o l a t i o n 121 ( i i ) Reaction with 2,4-Dinitrophenylhydrazine 124 ( i i i ) Aldehydo-hexoses by Oxidation of Hexitols 132 EXPERIMENTAL 136 General Considerations 136 Experimental Section A 138 3 , 4-Di-O-aeetyl-D-xylal 138 Bicobalt Octacarbonyl 139 Hydroxymethylation of 3 , 4-di-O-acetyl-D-xylal 140 Characterisation of Fractions I and II 142 Consumption of Periodate Ion 144 Enantiomeric 2-deoxy -3-0-(2-hydroxyethyl)-glycero- t e t r i t o l s ~ ~ 146 2-Deoxy - 3-0-(2-hydroxyethyl)-L-glycero-tetritol 148 v i i i Page Attempted Preparations of 2-deoxy - 3 , 4-di - 0-acetyl- D-xylopyranosyl cyanide 153 2 , 3-Di , - 0-acetyl-l , 5-anhydro - 4 , 6-dideoxy-D-arablno- and L-xylo-hexitols — 155 I d e n t i f i c a t i o n of Polyol Y from hydrogenolysis of methyl <x -D-glucopyranoside 160 Reaction of 3 , 4-di-O-acetyl-D-xylal with Carbon Monoxide and Deuterium -~ l 6 l Experimental Section B 164 3 , 4 , 6-Tri-O-acetyl-D-galactal 164 Reaction of 3 , 4 , 6 - t r i - O - a c e t y l - D - g a l a c t a l with Carbon Monoxide and HycTrogen — 166 Characterisation of Fractions A and B 167 2,6-Anhydro-3-deoxy-4,5-G-lsopropylidene-D-ta_lo_- h e p t i t o l — 169 236-Anhydro-3-deoxy-4,5-0-isopropylidene-1,7-di- 0-p-tolysulphonyl-D-talo"^heptitol 170 Reaction with sodium iodide 170 Consumption of Perlodate 171 Enantiomeric 2-deoxy - 3 - 0 -(l,3-dihydroxy - 2-propyl)- g l y c e r o - t e t r i t o l s 172 Attempted syntheses of o p t i c a l l y pure t e t r o l ethers 174 2-Deoxy -3-0_-(l,3-dihydroxy -2-propyl) -L-glycero- t e t r i t o l from 2 ,6-anhydro -3-deoxy-D-gfuco-heptitol 179 Experimental Section C 180 Hydroformylation of 3*4-di-£-aeetyl-D-xylal 180 Reaction with 2 , 4-dinitrophenylhydrazine 181 ix Page Fraction Y 182 1 ,5-Anhydro-4-deoxy-D-arabino-hexitol from Fraction Y = 185 Pi - 0-a c ety1-anhydrodeoxy-aldehydo-hexoses by oxicTation of di-O-acetyl-anhydrodeoxy-hexitols 185 REFERENCES ' 188 Figure Following page 1 82 2 86 3 97 4 113 ! - 1 - GENERAL INTRODUCTION In order to provide a background to subsequent d i s - cussion of the application of the oxo reaction to gl y c a l s , a b r i e f review w i l l be made of the oxo reaction of o l e f i n s i n general, and of the chemistry of the g l y c a l s , with p a r t i - cular emphasis i n each case on stereochemical aspects. The Oxo Reaction The reaction of o l e f i n s with carbon monoxide and hydrogen i n the presence of a cobalt catalyst i s commonly referred to as the oxo reaction, as early experiments using ethylene as substrate led to the formation of an appreciable quantity of d i e t h y l ketone. However, the reaction of ethy- lene i s not t y p i c a l , and i n general aldehydes or alcohols are the major products. The stoichiometry of aldehyde formation i s as shown i n equation l_s R.CH=CH.R + H 0 + CO > R.CH .CHR.CHO 1 2 2 . — 1 This reaction i s often referred to as hydroformylation , being formally equivalent to the addition of hydrogen and a formyl group at either end of the double bond. Alcohol formation i n the oxo reaction (equation 2) res u l t s from the further reduction of aldehydes formed by the hydroformylation reaction: R.CH2.CHR.CHO •+ ,6 > R.CHg.CHR.CHgOH 2 - 2 - As there i s no term i n general use to describe the o v e r a l l conversion of o l e f i n s to alcohols (equation 1_ + 2) i t i s proposed, f o r the sake of convenience, to coin the expression "hydrohydroxymethylation" to describe -the addition of hydrogen and a hydroxymethyl group to the double bond, and abbreviate t h i s to "hydroxymethylation" i n subsequent discussion. It i s noteworthy that t h i s second stage (equation 2_) from a mechanistic viewpoint i s c l o s e l y related to hydroformylation, and indeed requires the presence of an appreciable p a r t i a l 2 pressure of carbon monoxide , evidence that the active cata- l y s t i s a cobalt carbonyl. H i s t o r i c a l l y , the oxo reaction developed from the well known process f o r hydrocarbon synthesis discovered by Fischer and Tropsch , i n which hydrogen and carbon monoxide were passed over an iron catalyst under conditions of high temperature (400-450°) and moderate pressure. It was observed that small amounts of oxygen-containing product's 4 were often formed, and i n 1929 Smith, Hawk and Golden obtained an increased y i e l d of oxygenated material when ethylene was added to the carbon monoxide-hydrogen mixture before passing i t over a cobalt catalyst under Fischer- Tropsch conditions. Credit f o r the development of the oxo reaction as a commercial process goes larg e l y to Roelen and co-workers of Ruhrchemie A. G. i n Germany before and 5 during World War II . By changing the conditions of the - 3 - Pischer-Tropsch synthesis to give a higher pressure and a lower temperature, the reaction of ethylene with water gas (equal volumes of carbon monoxide and hydrogen) was made to y i e l d a product consisting of propionaldehyde together with some di e t h y l ketone, no hydrocarbons being formed. Following World War II the oxo synthesis attracted wide i n t e r e s t , i n p a r t i c u l a r because of i t s wide range of a p p l i - c a b i l i t y i n the conversion of o l e f i n s to aldehydes and a l - cohols, e s p e c i a l l y the l a t t e r . In general, temperatures between 7 5 ° and 2 0 0 ° , and pressures of synthesis gas from 100 to 300 atmospheres are employed, higher temperatures being usual when alcohols rather than aldehydes are the desired products. In step with the commercial development of the oxo" process, much fundamental work has been car r i e d out on t h i s and related reactions catalysed by the metal carbonyls, and 6,7,8 t h e i r chemistry has been well reviewed from time to time Despite the accumulation of a considerable amount of know- ledge of these reactions, many of the f i n e r points of the complex mechanisms are s t i l l not known with certainty. In the following pages a b r i e f survey w i l l be made of current views regarding the nature of the ca t a l y s t , and the probable ro l e which i t plays i n the conversion of o l e f i n s to aldehydes, and aldehydes to alcohols. - 4 - (i) Nature of the Catalyst In the early stages of development of the oxo synthesis i n Germany the catalyst used was the conventional Fischer- Tropsch surface c a t a l y s t , consisting of a mixture of metallic cobalt and Kieselguhr, together with small amounts of thorium and magnesium oxides. It became evident to Roelen and co- workers that a l l the components of the Fischer-Tropsch cata- l y s t with the exception of cobalt were superfluous, and that the active catalyst was probably a soluble carbonyl of cobalt formed by the i n s i t u reaction of the metal with synthesis gas. This was confirmed by subsequent investigations of the reaction on a laboratory s c a l e 1 ' 2 . Adkins and Kresk^, who introduced the use of preformed dicobalt octacarbonyl as c a t a l y s t , demonstrated that the hydroformylation reaction was i n s e n s i t i v e to sulphur poisoning, further evidence for the homogeneous nature of the c a t a l y s i s . The fact that the oxo reaction i s catalysed by a cobalt carbonyl contributes to the commercial importance of the process since the form i n which cobalt i s added i s not too important, as would be the case with a conventional s o l i d phase c a t a l y s t . In practice, crude organic s a l t s such as the octanoate or naphthenate are often used. The divalent cobaltous ion i s f i r s t reduced to the metal by hydrogen, E0 —> 2 H 4 + 2 e J Co + + + 2 e —> Co 3 - 5 - and the metal then reacts with carbon monoxide to form d i - cobalt octacarbonyl 2 Co + 8 CO Co.(CO) 4 2 8 ~ Hence, the presence of both carbon monoxide and hydrogen are required to convert cobalt s a l t s to the carbonyl. A considerable amount of evidenoe has accumulated which indicates that cobalt hydrotetracarbonyl, HCo(CO)^, rather than dicobalt octacarbonyl, i s e f f e c t i v e i n i n i t i a t i n g the oxo reaction. The hydrotetracarbonyl i s formed by reaction of hydrogen with dicobalt octacarbonyl C O 2 ( C 0 ) Q + H ^ 2 HCo(CO) 4 5 This step i s thus of fundamental importance i n that i t In- volves "the a c t i v a t i o n of molecular hydrogen 1 0, which i s trans- ferred from the gas to the l i q u i d phase. Orchin and co- workers1"1' have shown that the hydrotetracarbonyl i s present under oxo conditions i n the absence of o l e f l n j but when o l e f i n Is present no cobalt hydrotetracarbonyl i s detectable (as the Co(CO) a] ~) u n t i l hydroformyla-cobalt tetracarbonyl anion ^ N — , ^ t l o n of the o l e f i n i s complete, when an appreciable amount of the hydrotetracarbonyl again appears In the reaction mixture. ( i i ) Mechanism of the Oxo Reaction Evidence f o r the mechanism of the oxo reaction as - 6. - carried out under conditions of high temperature and pressure i s l a r g e l y speculative, and i s based on the r e s u l t s of re- actions between o l e f i n s and cobalt hydrotetracarbonyl at atmospheric pressure and temperature. Early mechanistic 1 12 13 theories ' ' did not take into account the now well-proven intermediacy of cobalt hydrotetracarbonyl, and no longer 12 appear to be relevant. An early finding of importance was that the rate of hydroformylatlon varied inversely with i n - crease i n the p a r t i a l pressure of carbon monoxide at constant" hydrogen pressure. The stoichiometry of the reaction of ole- f i n s with cobalt hydrotetracarbonyl and carbon monoxide at room temperature was investigated by Orchin and co-worker s ^ ' 1 ^ and with a moderate excess of o l e f i n (l-pentene) under 1 atmosphere of carbon monoxide was found to be 2 HCo(CO)^ + CO + o l e f i n — > Co 2(CO)g + aldehyde 6 The r e l a t i v e rates of reaction of various o l e f i n s under IS these conditions ^ was found to p a r a l l e l t h e i r rates of hydroformylatlon under oxo c o n d i t i o n s 1 ^ . The reaction of o l e f i n s with cobalt hydrotetracarbonyl at room temperature 17,18 and below was further explored by Heck and Breslow , on whose work current views regarding the mechanisms of the oxo reaction, separately discussed below f o r hydroformyla- t l o n and hydrogenation, are»largely based. (a) HydroformylatIon Subsequent to the generation of cobalt, hydrotetra- carbonyl by the hydrogenolysis of dicobalt octacarbonyl (equation 5 ), the conversion of o l e f i n s to aldehydes i s regarded as proceeding i n three d i s t i n c t stages; 1. forma- t i o n of a TT-complex between o l e f i n and cobalt hydro- carbonyl, which rearranges so that a carbon-metal slgma bond i s formed; 2 . i n s e r t i o n of carbon monoxide between metal and carbon, and 3 . hydrogenolysis of i'fche r e s u l t i n g complex to give an aldehyde. 17 1. Heck and Breslow consider that t h i s f i r s t stage involves at least three d i s t i n c t steps, as follows HCo(CO) 4 R.CH=CH.R + HCo(CO). HCo(CO) CO R.0H=CHR HOo(CO) 3 J RCH2.CHR.Co(C0). 8 RCH2.CHR.Co(CO)3 + CO <F=̂  RCHg.CHR.CotCO)^ 9 Their view that cobalt hydrotricarbonyl, rather than the hydrotetracarbonyl, i s the reactive species i s based on evidence that the formation of alkylcobalt tetracarbonyls i s i n h i b i t e d by carbon monoxide; more fundamentally, i n i t i a l complexing with o l e f i n would presumably require the p a r t i c i - pation of a co-ordinately-unsaturated carbonyl. - 8 - 2. Heck and Breslow ° found that methylcobalt t e t r a - carbonyl absorbed exactly one mole of carbon monoxide to give a product with a strong band i n the infrared at 1728 -1 cm , assigned to the acylcobalt linkage, R.CO.Co. The same inf r a r e d absorption at reduced i n t e n s i t y was also shown by solutions of alkylcobalt tetracarbonyls, i n d i c a t i n g that these complexes were i n equilibrium with acylcobalt t r l c a r b o n y l s , GO RCH2.CHR|.Co(CO)4 ̂  R.CH2CHR.CO.Co(CO)3 ^ RCHg.CHR.CO.Co(CO)4 10 Evidence has been obtained with analogous complexes of 19 14 manganese , using C - l a b e l l e d carbon monoxide, which i n - dicates that the inserted carbonyl was o r i g i n a l l y bonded to the metal. The i n s e r t i o n reaction i s thus e s s e n t i a l l y the migration of an a l k y l group from metal to the carbon atom of a carbonyl ligand. 3. The reaction of acetylcobalt tetracarbonyl with cobalt hydrotetracarbonyl under room temperature conditions afforded acetaldehyde and dlcoba.lt octacarbonyl i n good y i e l d 1 8 CH3CO.Co(CO)^ + HCo(CO) 4 > CH^HO + Co 2(C0) Q 11 This reaction i s not, however, considered to operate under oxo conditions as, despite the fact that acetylcobalt - 9 - tetracarbonyl i s also reduced to aldehyde by hydrogen under pressure, the reaction i s completely i n h i b i t e d by carbon monoxide. To account f o r the f i n a l stage of hydroformyla- t i o n , Heck and Breslow suggest the intermediacy of eo- ordinately-unsaturated acylcobalt t r i c a r b o n y l s , which are reduced to alddydes by hydrogen or converted to unreactive tetracarboriyls by carbon monoxide. The well-known adverse eff e c t of carbon monoxide on the course of the oxo reaction can therefore be attributed to t h i s competition. RCH 0.CHR.CO.Co(CO)„ + Ho — » RCH^.CHR.CHO +HCo(CO)o 12 2 3 d 2 5 — CO If S HCo(CO)o + CO ̂  HCo(CO)u RCHg.CHR.CO.CoCCO)^ 3 (b) Hydrogenation A scheme which i s analogous to that described above 1*7 l ft 20 fo r hydroformylation '* has been proposed by Marko for the subsequent hydrogenation of aldehydes to alcohols under oxo conditions (equation 2_) . Go-ordinately-unsaturated carbonyls are considered to be the reactive intermediates; thus, cobalt hydrotrlcarbonyl forms a 1\ -complex with the aldehyde, which rearranges to an alkoxycobalt t r i c a r b o n y l (equation 1 3 ). Marko suggests that t h i s Is then hydro- genolysed by molecular hydrogen to give the alcohol, or i s competed f o r by carbon monoxide, giving r i s e to an unreactive tetracarbonyl - 10 - R.CHO + HCo(CO) 3 ,H RC=0 v HCo(CO) R.CH20.Co(C0) R.CH 20,Co(CO) 3 + H 2 >R.CH20H'+ HCo(C0) 3 13 14 CO R.CH2O.Co(CO)4 21 22 The views of Aldridge and Jonassen ' d i f f e r from those of other workers i n that they regard hydroformylatlon, and hydrogenation of aldehydes, to be heterogeneously catalysed reactions i n that the i n s i t u formation of cobalt hydrotetracarbonyl by the hydrogenolysis of dicobalt octacarbonyl (equation 5) i s catalysed by cobalt metal, which they consider to be present i n equilibrium with the soluble octacarbonyl (equation 4). Their concept of the subsequent stages i s , however, fundamentally si m i l a r to 17 18 20 those of Heck and Breslow 1' and Marko ( i i i ) E f f e c t of O l e f i n Structure on Hydroformylatlon ( a) E f f e c t on Rate of Reaction A l l simple o l e f i n s have been found to undergo the oxo reaction, although the rate of reaction i s observed to be highly dependent on the structure of the o l e f i n . Workers l 6 at the U.S. Bureau of Mines have made a comprehensive study of the influence of structure on the rate of hydro- formylatlon at 1 1 0 ° , using 26 o l e f i n i c hydrocarbons, and - 11 - a 5 0 - f o l d v a r i a t i o n was observed between the fastest and slowest rates. The r e s u l t s i n a l l cases appear to demonstrate a clear r e l a t i o n s h i p between reaction rate and degree of s t e r i c hindrance about the double bond, the l a t t e r factor presumably r e f l e c t i n g the ease of formation of an intermediate complex with cobalt hydrotricarbonyl. Straight chain terminal o l e f i n s were observed to react most r e a d i l y , with l i t t l e de- crease i n rate with increasing chain length. The rate for straight chain i n t e r n a l o l e f i n s was about one t h i r d that for the terminal o l e f i n s ; however, the exact position of the double bond, providing i t was i n t e r n a l , had l i t t l e influence on rate. Branching of the carbon chain always resulted i n a decrease i n reaction rate, even for o l e f i n s having a single methyl group remote from the double bond, as i n 4-methyl-1- pentene. The presence of a methyl group at one of the carbon atoms of the double bond reduced the rate 1 0-fold, f o r example i n going from 1-pentene to 2-methyl-l-pentene>. The slowest rates were observed with branched i n t e r n a l o l e f i n s : thus the rate of hydroformylation of 2,3-dimethyl-2-t|utene, i n which the double bond i s completely substituted by methyl groups, was approximately ^-/^Qth that of an unbranched terminal o l e f i n . The rates observed f o r c y c l i c o l e f i n s are of inte r e s t i n that, whereas cyclopentene and cyclo- heptene reacted faster than the corresponding a c y c l i c i n - t e r n a l o l e f i n s , the reaction rate for cyclohexene was - 12 - appreciably slower. This observation has been explained 0 on the basis that both cyclopentene and cycloheptene are i n a more highly strained state ( s t r a i n energies of 4.4 and 4.1 kcal/mole respectively r e l a t i v e to cyclohexene). 23 Traynham has pointed out that i n ajll known cases the more strained i s a c y c l i c structure, the more reactive i t i s i n electron donating r o l e s . One might therefore expect the TT electrons of cyclohexene to be less available f o r donation to the vacant d o r b i t a l s of cobalt i n complex formation. (b) E f f e c t on Mode of Addition In the hydroformylatlon of unsymmetrical o l e f i n s , available evidence indicates that the formyl group adds to the least hindered side of the double bond under normal conditions of high temperature and pressure, (although t h i s i s not necessarily the case for reactions with cobalt hydro- tetracarbonyl at room temperature1"^) J thus o l e f i n s having 9 H3 the structure R— CH=CH 2 give predominantly the terminal aldehyde. For example, the major product from the hydro- .formylation of isobutylene i s isovaleraldehyde, together oil with a small proportion of trimethylacetaldehyde^. \ 3 \ 3 j}==CH2 + CO + H 2 > J3H.CH2.CHO+ (CH3)3C.CHO CH 3 . CH 3 - 13 - Of p a r t i c u l a r interest i s the d i s t r i b u t i o n of products obtained from the application of the oxo reaction to c y c l i c v i n y l i c ethers, as these compounds are s t r u c t u r a l l y related to the g l y c a l s . The hydroxymethylation of 2 , 3-dihydro - 4 H - pyran (l) and cert a i n of i t s derivatives has been i n v e s t i - gated recently by Falbe and Korte 2^. The reactions were performed under conditions ( 190° , 300 atmospheres of syn- thesis gas (1*1))-' leading to the complete conversion of aldehydes to alcohols, thereby f a c i l i t a t i n g the i d e n t i f i - cation of reaction products. Reaction of 2 , 3-dihydro - 4 H - pyran (l) with carbon monoxide and hydrogen resulted pre- dominantly i n attachment of the hydroxymethyl group to the side of the double bond adjacent to the ri n g oxygen to give 2-hydroxymethyl-tetrahydropyran (2) i n 78$ y i e l d * i n a d d i t i o n ^ a small proportion of 3-hydroxymethyli- tetrahydropyran (3) was i s o l a t e d , together with a smaller amount, of tetrahydropyran (4) r e s u l t i n g from hydrogenation of the double bond CO/H2 COp^CO^; / ^ C H O H 8 N ) ^ C H 2 0 H X T (2) (3) 2 + (4) 3$ 16 Under s i m i l a r conditions 2-hydroxymethyl-2,3-dihydro- 4H-pyran (5) gave exclusively 2 ,6-bis-hydroxymethyl- tetrahydropyran (6) - 14 - HObLC'Xo (5) (6) 17 When the carbon atom of the double bond adjacent to the r i n g oxygen bore a methyl substituent, as i n 2 , 6-dimethyl- 2,3-dihydro-4H-pyran ( 7 ) , then the d i r e c t i o n of addition was reversed and 3-hydroxymethyl-2,6-dimethyl-tetrahydropyran (8) was i s o l a t e d , but i n lower y i e l d (7) C0/H 2 Co 2(C0) 8 > '2 "3 220 18 (8) Comparable res u l t s have previously been obtained by the hydroxymethylation of furan ( 9 ) . This compound reacted as a t y p i c a l conjugated diene, i n that one double bond was hydrogenated while the other underwent the normal 26 reaction with carbon monoxide and hydrogen ; addition of the hydroxymethyl group occurred at C - 2 , 2-tetrahydrofurfuryl alcohol (10) being i s o l a t e d i n 35$ y i e l d . + CO + 3H, C O 2 ( C 0 ) Q 19 (9) do) - 15 - However, when both positions adjacent to the r i n g oxygen were blocked, hydroxymethylation occurred at the al t e r n a t i v e s i t e . Thus, 2 ,5-ditrtethylfuran ( n ) gave 2 , 5-dimethyl - 3 - tetrahydrofurfuryl alcohol (12) 27 (11) H. (12) '3 20 A reaction which c l o s e l y resembles hydroformylation i s the hydrocarboxylation of o l e f i n s i n the presence of n i c k e l carbonyl + ROH H — H Ni(C0) 4 .28 ~ H •COOR • 21 -COOH — B i r d and co-workers found recently that strained o l e f i n s such as norbornene (13) underwent hydrocarboxylation at atmospheric pressure and temperature. By the use of deuterated solvents (deuterium oxide, deuterioethanol and deuterioacetic acid) It was shown that (13) was converted into 3-exo-deuterio-bIcyclo- j ^ 2 . 2 .lj -heptane-2-exo-carboxylic acid ( 1 4 ) . These workers suggest that, In hydrocarboxylation, 22 COOH - 16 - els addition to o l e f i n s from the least hindered side i s probably general. Two groups of workers 2^ 3 ° ^av.Q investigated the pre- paration of 6-methyl steroids by the application of the oxo reaction to steroids having a double bond i n the A ^ - p o s i t l o n (15). The product i n each case was the 6-<x-hydroxymethyl- a l l o s t e r o i d ( 16) , and i t was concluded that "the oxo reaction (15) ( i 6 r CH 2 OH 23 under hydroxymethylation conditions appears to take place by a c i s addition" . The Glycals (i) Structure of the Glycals 31 <• The g l y c a l s , discovered i n 1913 by Fischer and Zaeh , owe t h e i r name to Fischer's observation that the f i r s t (impure) preparations of D-glucal gave c h a r a c t e r i s t i c alde- hyde reactions. When i t was l a t e r r e a l i z e d that pure gly- cals are not aldehydie compounds, the name was fi r m l y established. A proposal to systematise g l y c a l nomenclature, - 17 - at present under o f f i c i a l consideration (cf reference 3 2 ) , regards the glycals as 1 , 5-anhydro-derivatives of 1-enolsj thus 3 , 4 , 6-tri-O-acetyl-D-galactal (40) would be renamed 3 , 4 , 6-trl - 0-acetyl-l , 5-anhydro - 2-deoxy-D-lyxo_-hexose-l-enol, Glycals are c y c l i c carbohydrates characterised by the presence of a CH=CH linkage between carbon atoms 1 and 2 . They can be regarded formally as derived from the corre- sponding aldose by removal of two hydroxyl groups from ad- jacent carbon atoms, (e.g. D-galactal (19) from D-galactose (18)), and i n t h i s respect d i f f e r from the other well known class of unsaturated carbohydrates, the glycoseens, which s t r u c t u r a l l y are derived from the corresponding aldose by -removal of a molecule of water. The l a t t e r group includes the so-called 2-hydroxyglucals (e.g. 2-hydroxy-D-ga1acta1 ( 1 7 ) ) , or 1 ,2-glycoseens, as well as the 5 ,6-glycoseens which have an exocyclic double bond between carbon atoms 5 and 6. , CH 2 OH -H20 QH^OH — ( X C H 2 0 H OH >H ;OH (18) (19) 24 Aldoses which are epimeric at C - 2 , such as D-glucose and D-raannose, give the same g l y c a l , as asymmetry at C-2 - 18 - (and C-l) i s l o s t on formation of the double bond; the name of the r e s u l t i n g g l y c a l i s usually derived from that of the more common of the two parent aldoses. Thus the eight stereoisomerie hexoses of the D-series can give r i s e to four "hexals"j D-glucal, D-galactal, D - a l l a l and D-gulal (no preparations of the l a t t e r two are known), and two 'pentals", (D-xylal and D-arabinal) are possible from the four D-pentoses. With the exception of a novel "furanal" ~~ 32 reported recently^ , a l l the known glycals are six-membered ring structures derived from pyranose sugars. The structure of the best known g l y c a l , 3 * 4 , 6 - t r i - O-acetyl-D-glucal (20 a) was worked out at an early staged* 3^ The presence of a double bond was demonstrated by the addition of two atoms of bromine or hydrogen; i t s p o s i t i o n , between carbon atoms 1 and 2, was shown by the fact that cleavage with ozone l i b e r a t e d D-arabinose, formation of the l a t t e r pentose also providing evidence that the configurations of carbon atoms 3» 4 and 5 are unchanged from those of ID- glucose. The structures of other g l y c a l s have been proved CHpR / — 0 (20a) R=Ac (20b) R = H - 19 - i n a s i m i l a r manner. The presence of the double bond i n the six-membered rin g of glycals results i n carbon atoms 1, 2 and 3 and the ri n g oxygen atom being constrained into the so-called half-chaix| conformation of the cyelohexene r i n g . The presence of substituents on the r i n g permits the p o s s i b i l i t y of two hailf-chair conformat^rm^or the g l y c a l s , thus, i n the case of one conformation (21 a) for 3,4,6-tri-O-acetyl-D-glucal, substituents at G-3, C-4 and C-5 are i n equatorial or pseudo-equatorial positions, whereas i n the a l t e r n a t i v e conformation (21 b) these sub- stituents are In the less stable a x i a l or pseudo-axial orientations. That 3,4,6-trI-O-acetyl-D-glucal adopts the (21a) CHfAc OAc (21b) more stable h a l f - c h a i r conformation (21 a) i n solution 35 was confirmed recently by H a l l and Johnson . Using proton magnetic resonance spectroscopy at high radio frequency (100 Me/s) these workers were able to measure the chemical s h i f t of each proton, and the coupling constants between adjacent protons, cert a i n of t h e i r assignments being confirmed - 20 - by double resonance experiments. I t was then possible to rela t e the coupling constants of adjacent r i n g protons to t h e i r dihedral angle by app l i c a t i o n of Karplus' equation-^, and thus ar r i v e at the approximate geometry of the r i n g . The r e s u l t s showed that the dihedral angle between C-H(3) and C-H(4) i s approximately 140°, and that between C-H(4) and C-H(5) i s about 1 5 0 ° , and therefore confirmed that the g l y c a l adopts the conformation (21 a) i n solut i o n . As the l a t t e r dihedral angle was- less than 180°, some " f l a t t e n i n g " of the ri n g was indicated. No comparable measurements have yet been made of the conformations of the other available g l y c a l s ; c e r t a i n l y i t i s reasonable to predict that the ha l f - c h a i r conformation of 3 ,4-dl-Oj-acetyl-D-xylal (22) would be analogous to that found for 3 , 4 , 6-tri-O-acetyl- D-glucal. ( i i ) Preparation of the Glycals Despite the fact that over 50 years have elapsed since Fischer and Zach f i r s t synthesised 3 i 4 i 6-tri - 0-acetyl-D- g l u c a l , t h e i r procedure, with minor modifications designed (22) - 21 - to Improve y i e l d s , i s s t i l l the only one of significance f o r the preparation of g l y c a l s . I t appears probable^ 7 that Fischer's o r i g i n a l intention on treating 2 , 3 , 4 , 6 - t e t r a - 0 - acetyl-cx-D-glucosyl bromide (23) with zinc and acetic acid was reductive dehalogenation to a f f o r d a derivative of 1 ,5- anhydro-D-glucitoi; i n actual fact a good y i e l d of 3 , 4 , 6 - tri-O-aeetyl-D-glucal (20 a) was obtained, AcO Zn/AcOH 25 (20a) (23) A mechanism to explain g l y c a l formation has been suggested by P r i n s ^ 8 , and discussed by Overend and Stacey3°\ The carbonium ion intermediate (25) r e s u l t i n g from i o n i s a t i o n of the acetobromoaldose (24) can either react with solvent to give the acetylated aldose ( 2 6 ) , or can acquire two electrons from the metal to furnish the carbanion ( 2 7 ) , which affords the g l y c a l by elimination of an acetoxy group from C - 2 . It i s also conceivable that the carbanion may acquire a proton to give the 1 , 5-anhydroalditol (29)i although the l a t t e r do not appear to have been detected as products of t h i s reaction. - 22 - H-C-Br H-j-OAc (24) 9 I 2 H-C-OAc I (29) 6 - f t H- £ 7 * H-O-OAc (25) 2e ^ -C^OAc 0 (27) * H-C-OAc 1 • ? • (26) R=OH or OAc 26 HC . o I (28) Because of the Importance of glycals as intermediates i n carbohydrate chemistry, i n p a r t i c u l a r i n the preparation of 2-deoxy s u g a r s ^ , various modifications to Fischer's o r i g i n a l procedure have been introduced with the object of improving the o v e r a l l y i e l d s of the sequence: a l d o s e — > acetobromoaldose — > g l y c a l — > 2-deoxy-aldose. A standard method fo r the preparation of acetobromoaldbses was passage of hydrogen bromide through a solution or sus- pension of the aldose i n acetic anhydride. A s i g n i f i c a n t improvement over the dir e c t use of hydrogen bromide was introduced by Barczai-Martos and Korosky , who generated the gas iri s i t u by adding phosphorus tribromide and water, or more simply phosphorus,•.•\toi?6mlne and water>- to -a solution of the f u l l y - a c e t y l a t e d aldose In acetic anhydride. Good y i e l d s of several c r y s t a l l i n e acetobromoaldoses were ob- tained i n t h i s way, without Intermediate i s o l a t i o n of the f u l l y - a c e t y l a t e d precursors. - 23 - Various modifications have been made to Fischer's o r i g i n a l procedure for reduction of the acetobromo-sugar 4 l with zinc and acetic a c i d . Deriaz and co-workers i n - troduced the use of c h l o r o p l a t i n i c acid which, added at i n t e r v a l s to the reduction mixture, maintained a vigorous reaction and enabled lower temperatures to be employed. S i g n i f i c a n t improvements i n the y i e l d s of 3 ,4-di-0-a cety1-D- 4T I — lip arabinal and 3,4,6-tri-O-acetyl-D-galactal were attained i n t h i s way. I s e l i n and Reichstein^^ added sodium acetate to buffer the zinc-acetic acid mixture and to remove hydrogen bromide formed during the course of the reaction, and activated the zinc by addition of copper as the sulphate. The most con- venient general procedure f o r g l y c a l preparation i s that of 44 45 H e l f e r i c h and co-workers ' , who combined the advantages of the acetobromoaldose preparation of Barczai-Martos and 40 Korosky , and the reduction procedure of I s e l i n and 43 Reichstein * the conversion of aldose to acetylated g l y c a l i s c a r r i e d through i n one stage without i s o l a t i o n of i n t e r - mediates. ( i i i ) Reactions of the Glycals P r a c t i c a l l y a l l the reactions of the glycals have t h e i r counterpart i n well known reactions of o l e f i n s i n general, involving addition across the double bond. In the majority of cases the d i r e c t i o n of addition i s Influenced - 24 - by the proximity of the rin g oxygen atom. Glycals react as t y p i c a l v i n y l ethers, and frequently exhibit an in t e r e s t i n g resemblance to the six-membered c y c l i c v i n y l ether, 2 , 3-dihydro- 4H-pyran ( l ) . Typical e l e c t r o p h i l i c addition to a v i n y l ether r e s u l t s i n addition of the electr o p h i l e to the carbon of the double bond (i to the ether oxygen, as the r e s u l t i n g cation i s s t a b i l i s e d by the el e c t r o n - r i c h oxygen. Acid-catalysed formation of the well known tetrahydropyranyl ethers (30) 46 thus proceeds as follows The r i n g oxygen of glycals Is seen to exert a s i m i l a r d i r e c t - ing Influence on the course of additions to the double bond; moreover, s t e r i c factors due to substituents at C-3 appear to have a marked eff e c t i n determining the d i s t r i b u t i o n of isomeric products. Examples can be taken from a variety of addition reactions of glycals to i l l u s t r a t e these points. Ionic additions of reagent of the general form1 HX (where X =0H, alkoxy, halogen) to the double bond of glycals are well known reactions f o r the preparation of 2-deoxy aldoses - 25 - and t h e i r d e r i v a t i v e s . The acid-catalysed addition of water to afford 2-deoxy sugars has been studied i n d e t a l l 3 9 ; t y p i c a l l y the conversion i s effected i n cold, d i l u t e sulphuric acid s o l u t i o n . I t was considered by I s b e l l and Plgman^ 7 that under these conditions 2-deoxy-D-galactose was formed from D-galactal (19) via an Intermediate sulphate ester which was subsequently hydrolysed on heating with barium carbonate J 48 however, t h i s was disproved by Overend and co-workers . An alt e r n a t i v e 'mechanism involving proton-catalysed opening of 4Q the oxygen bridge has been suggested •7. The accepted mechan- 46 ism of e l e c t r o p h l l i c addition to v i n y l ethers , exemplified by the formation of tetrahydropyranyl ethers (30) (equation 27) would appear to be applicable to t h i s and other ionic addi- tions of HX-type reagents to g l y c a l s . Gis Additions to the Double Bond The reaction of o l e f i n s with n i t r c s y l chloride, with the formation of nitroso-derivatives, has been c o n s i d e r e d ^ to proceed by an ionic mechanism, with the nitroso group entering into combination as an e l e c t r o p h l l i c fragment N|Q+, and the chlorine as nucleophilic C l ~ . Meinwald and co- 51 workers have obtained experimental evidence from a study of the reactions of norbornene and norbornadiene with n i t r o - s y l chloride which casts doubt on the ionic mechanism (apparent c i s addition, l a c k of incorporation of nucleophilic - 26 - solvent). They suggest a four-centre mechanism of addition with l i t t l e or no carbonium ion character developing i n the t r a n s i t i o n state 8- Serfontein and co-workers52 have obtained highly c r y s t a l l i n e adducts by the reaction of n i t r o s y l chloride with various acetylated g l y c a l s . Structural investigations by proton magnetic resonance spectroscopy confirmed that addition to the double bond was c i s . The fact that from 3 , 4 , 6 - t r i - O-acetyl-D-glucal ( 2 0 a ) , 3 , 4 , 6 - t r i - 0 - a c e t y l - 2 - d e o x y - 2 - nitroso-oc-D-glucopyranosyl chloride (31) was obtained, whereas the (i -D-arabinopyranosyl chloride (33) resulted from the addition of n i t r o s y l chloride to 3 , 4-di-O-acetyl- D-arabinal (32), would seem to indicate that the C-3 acetoxy group influences the d i r e c t i o n of approach of n i t r o s y l chloride to the double bond. AcO CHpAc )Ac (20a) N0C1 CH20Ac O OAc 28 AcON Jt\ N=0 (3D - 27 - N0C1 AcO' AcO 29 OAc (32) The light-catalysed reaction of o l e f i n s with 53 phenanthraquinone to give 1,4-dioxane derivatives^ was applied to the glycals by H e l f e r i c h r ^ . When 3 , 4 , 6 - t r i - O - acetyl-D-gli|cal (20 a) was thus reacted with phenanthra- quinone and the r e s u l t i n g adduct deacetylated, phenenthrene- hydroquinone-D-glucoside anhydride (34) was obtained, from which D-glueose (35) was l i b e r a t e d on ozonolysis. The fact that no D-mafnnose was obtained from these reactions indicates that addition of phenanthraquinone to the double bond of the g l y c a l took place exclusively from the least hindered side CH 20H AcO S i m i l a r l y , c i s hydroxylatlon of the double bond of D-glu'cal (20 b) and 3 , 4 , 6-tri-O-acetyl-D-glucal (20 a) by - 28 - osmium tetroxide, proceeding through the c y c l i c diester of osmic acid (equation 31) resulted i n predominant formation of the c y c l i c intermediate at the least hindered side of the double bond, and i n both cases more D-glucose than 55 — D-mannose was obtained on hydrolysis"^ (b) Trans Additions to the Double Bond Although less relevant i n considering possible s t e r l e e f f e c t s l i k e l y to be operative i n the hydroformylatlon of across the double bond of acetylated glycals are known i n which a s i m i l a r d i r e c t i n g influence appears to be exerted by the C-3 acetoxy substituent. The reactions i n question are considered to proceed through c y c l i c intermediates which open by rearward attack of a nucleophile to give a trans product. I t i s known that i n c y c l i c systems the preferred s t e r i c course of trans addition i s such as to favour the 56 d i a x i a l product . Thus, In the case of the g l y c a l s , where nucleophilic addition i s at C - l , trans d i a x i a l opening of a c y c l i c intermediate requires that the l a t t e r be i n a HO. 31 g l y c a l s , a number of reactions involving tlyglns addition - 29 - ^ - o r i e n t a t i o n (above the plane of the ring) ( 3 6 ) . Con- versely, a trans-dlequatorial product must arise from a c y c l i c intermediate (37) which i s oriented below the plane of the pyranoid ri n g (37) The addition of mercuric s a l t s to o l e f i n s i s usually considered to proceed via a mercurinium ion (e.g. ( 3 8 ) ) , which reacts with nucleophilic solvent to give a trans p r o d u c t ^ . From 3 , 4 , 6-tri-O-acetyl-D-glucal (20 a), 5 8 — Manolopoulos and co-workers^ obtained, on reaction with mercuric acetate i n methanol, a c r y s t a l l i n e compound i d e n t i - f i e d as methyl 3 , 4 , 6-tri - 0-acetyl - 2-acetoxymercuri - 2-deoxy- &_D-glucopyranoside ( 3 9 ) , the trans-di'^'quatorial product - 30 - These workers believe that the bulky C-3 acetoxy group tends to s h i e l d the double bond from attack at the upper (^) side, and the cx_mercurinium ion (38) becomes important. Rearward approach of methanol to C-l then leads to (39). Other reactions of 3 ,4,6-tri-O-acetyl-D-glucal proceeding through analogous mechanism, i n which the forma- t i o n of an equal or major proportion of the k i n e t i c a l l y less favoured product having the D-gluco-configuration (C-2 equatorial substituent) indicates the importance of a c y c l i c intermediate at the least hindered (oc) side of the r i n g , are with perbenzoic acid (via 1,2-epoxide)-*0^ 60 with bromine (via bromonium ion) , and with iodine and 6l s i l v e r benzoate (via iodiniura ion) The Oxo Reaction of Glycal3 The reaction of glycals with carbon monoxide and hydrogen under oxo conditions was f i r s t explored i n 1956 62,63,64 by Rosenthal and co-workers . Glycals studied were 3,4,6-tri-0-acetyl-D-galactal (40) and 3,4,6-tri-O-acetyl- D-glucal (20 a); i n each case saturated seven carbon com- pounds were obtained and characterized. CHpAc ; AcO, OAc (40) - 31 - The oxo reaction of glycals therefore represented an additional 62 method for lengthening the carbon chain of carbohydrates . The products obtained from these reactions were acetylated alcohols, rather than aldehydes, and undoubtedly arose by the i n i t i a l hydroformylation of the g l y c a l double bond, followed by reduction of the formyl group to hydroxymethyl (equations IL and 2 ). The structures of the products thus obtained were not established, but i n the case of the reac- t i o n of 3,4,6-tri-o_-acetyl-D-galaetal i t was thought that hydroxymethylation had occurred at C-2 of the g l y c a l , to 63 give a branched-chain carbohydrate ; t h i s g l y c a l also appeared to be unique i n that only one product was iso l a t e d from i t s reaction with carbon monoxide and hydrogen. e a r l i e r investigations, to which further reference w i l l be made i n subsequent discussion, to a study of the oxo re- action of the pental, 3,4-di-0-acetyl-p_-xylal ( 2 2 ) , and also describes a further investigation of the reaction of 3,4,6- tri-O-acetyl-D-galactal (40). H-work discussed i n t h i s thesis extends these - 32 - DISCUSSION A. Hydroxymethylation of 3,4-Di-O-acetyl-D-xylal Previous work^3,64 h a s demonstrated that the oxo reaction can be applied successfully to the acetylated hexals (20 a) and(4o), the major products of the reactions being acetylated alcohols having one more carbon atom than the s t a r t i n g material; thus the reactions appeared to follow the expected course and a hydroxy methyl.'', group was added to one side of the double bond. It could be a n t i c i - pated that application of the same reaction to the acety- lated pentals would therefore r e s u l t In the formation of di-0-acetyl derivatives of six carbon compounds. The ob- ject of the work described i n t h i s section was to i n v e s t i - gate the structures of products formed by the reaction of 3,4-di-O-acetyl-D-xylal (22) with carbon monoxide and hydrogen under hydroxymethylation conditions, whereby, as a resu l t of the absorption of 3 moles of synthesis gas per mole of substrate, the anticipated products are alcohols, rather than aldehydes -CH=CH- + ^2H g + CO, -> -CH2-CH-CH20H 34 3 moles In a l a t e r section (C) experiments w i l l be described i n which i t was attempted to terminate the reaction at the aldehyde stage (equation l ) , ' a t the point where 1 mole of - 33 - g l y c a l had absorbed 2 moles of synthesis gas. (i) Reactants and Reaction Conditions (a) 3 , 4-Di-O-acetyl-D-xylal (22) 3 , 4-Di-O^acetyl-D-xylal (22) was f i r s t prepared "~ 65 i n 1929 by Levene and Mori , as an intermediate i n the synthesis of 2-deoxy- -D-xylose, using an adaptation of ~ 31 the o r i g i n a l procedure of Fischer and Zaeh^ . Reduction of 2 , 3 > 4-tri - 0-acetyl-oc-D-xylopyranosyl bromide (43) with zinc and 50$ acetic acid gave (22) i n 60$ y i e l d . E s s e n t i a l l y the same procedure has recently been given as a standard OAC ;H,CAC -> AcO CAc OAc (22) , 35 Br CAc (42) (43) 66 method fo r the preparation of t h i s g l y c a l . Overend and 67 • co-workers ' have pointed out the necessity of employing low temperatures (-5 to -10 ) f o r the conversion of aceto- bromoaldoses to g l y c a l s , p a r t i c u l a r l y i n the pentose series, i n order to minimise the simultaneous formation of saturated products ((25) — > ( 2 6 ) ) . In view of t h i s , the y i e l d of 68 (22) reported by Gakhokidze (80$ from (43)) was remarkably high as the reaction was performed at room temperature. _ 34 - Attempts to repeat th i s preparation, however, have resulted only i n the i s o l a t i o n of 2 , 3 , 4-tri-O-acetyl-D-xylose i n ••M' 6Q rather good y i e l d y , In t h i s work, the procedure of H e l f e r i c h and co- 45 workers , i n s l i g h t l y modified form, was used to prepare 3 , 4-di-O-acetyl-D-xylal. Conversion of ( 4 l ) to (22) was thus car r i e d through without i s o l a t i o n of the intermediate compounds (42) and (43). Consistently s a t i s f a c t o r y re- sults were obtained using t h i s procedure providing the following precautions were observed; (a) The entire preparation was car r i e d through as quickly as possible, e s p e c i a l l y the f i n a l stage (43) > ( 2 2 ) . (b) The temperature of the zinc-acetic acid reduction mixture was not allowed to exceed -10°; addition of s o l i d carbon dioxide was e f f e c t i v e i n maintaining a low tempera- ture, both during the course of the reaction and the sub- sequent f i l t r a t i o n . (c) Normally, removal of zinc by f i l t r a t i o n i s extremely slow, and as some r i s e i n temperature Is inevitable during t h i s stage, undue delay i s l i k e l y to lead to reduced y i e l d s . It was found that f i l t r a t i o n could be greatly speeded by adding C e l i t e to the reaction mixture, and spreading a layer of C e l i t e on the ijjLlter paper before f i l t r a t i o n . • (c) The crude syrupy product obtained by chloroform ex- t r a c t i o n of the f i l t r a t e was d i s t i l l e d without delay. The - 35 - product thus obtained c r y s t a l l i s e d spontaneously, and suffered no apparent decomposition when stored f o r several months i n the r e f r i g e r a t o r . T y p i c a l l y , by t h i s procedure, 3>4-di -0-acetyl-D- x y l a l (22) was obtained from D-xylose (Hi) i n an o v e r a l l y i e l d of about 60$. The purity of the product obtained by d i s t i l l a t i o n was r e a d i l y demonstrated by th i n layer chromatography, acetylated glycals being considerably more mobile on s i l i c a gel than other components present i n a crude preparation. The most c h a r a c t e r i s t i c region of i n f r a r e d absorption of (22) i s a sharp, f a i r l y strong band at 1640 cm - 1, assignable to the C=C stretching v i b r a - t i o n . The presence or absence of t h i s band i s therefore of use i n following the course of reactions involving additions to the double bond. (b) Reaction Conditions Suitable conditions f o r the reaction of acetylated glycals with carbon monoxide, hydrogen and dicobalt octa- carbonyl have been well established as a r e s u l t of pre- vious work^2*. The early experiments of Adkins and Kresk^ suitably adapted the oxo reaction to a laboratory scale, and i n general l i t t l e deviation from these conditions i s observed i n the experiments of subsequent workers. How- ever, i n the case of the reaction of acetylated g l y c a l s , - 36 - lower temperatures than are often customsfry have been used, o Whereas at 130 glycals are la r g e l y converted to alcohols by hydroxymethylation (equation 2), with the majority of o l e f i n s aldehydes are formed predominantly at t h i s tempera- ture, and appreciable alcohol formation i s obtained only when the temperature i s raised by 5©° or more. The most rapid reaction rate i s obtained when the r a t i o of hydrogen 12 to carbon monoxide i s high ; the l i m i t i n g factor would appear to be the requirement that the p a r t i a l pressure of carbon monoxide i s s u f f i c i e n t l y high to prevent decomposition 2 of the catalyst . In practice, 3,4 , 6-tri-O-acetyl-D-galactai (40) has been observed to react normally with carbon monoxide and hydrogen when the i n i t i a l p a r t i a l pressures of the two gases were 200 and 1700 p . s . i . r e s p e c t i v e l y ^ . For the hydroxymethylation of 3*4-di - 0-acetyl-D-xylal, a carbon monoxide to hydrogen r a t i o of 1^5 was employed, at a t o t a l i n i t i a l pressure of about 3000 p . s . i . (c) Removal of Catalyst Wender^® has described various methods f o r the removal of dicobalt octacarbonyl catalyst from oxo reaction products. When the i s o l a t i o n of aldehydes i s not required, the preferred method i s to replace unreacted synthesis gas with hydrogen under pressure, when, on heating, the octa- carbonyl Is decomposed to metallic cobalt. A l t e r n a t i v e l y , - 37 - dicobalt octacarbonyl may be destroyed by heating the mixture on a steam bath, or shaking with d i l u t e sulphuric?K acid solution, u n t i l carbon monoxide i s no longer evolved. In working with the products derived from the acetylated glycals we have found the most convenient? method fo r separating reaction products from catalyst Ms by f i l t r a t i o n through P l o r i s i l (a synthetic magnesia-silica gel absorbent); catalyst i s eluted with petroleum ether, and reaction pro- ducts are subsequently eluted with a more polar solvent, such as a 9*1 (v/v) mixture of benzene and ethanol. ( i i ) Fractionation and Characterisation of Reaction Products (a) Chromatographic Separation of Products Evidence that the anticipated hydroxymethylation of the double bond of 3,4-di-O-acetyl-D-xylal had occurred was provided by the i n f r a r e d spectrum of the c a t a l y s t - f r e e product i s o l a t e d from the reaction: the strong absorption at 1640 cm"1 c h a r a c t e r i s t i c of the g l y c a l double bond had disappeared, and a band of moderate i n t e n s i t y had appeared i n the 3400 cm - 1 region (OH stretching;) . Thill layer chromatography showed the presence of a mixture with two major components which were not well resolved, and indicated that l i t t l e would be gained i n attempting to fractionate the mixture as such. Attempted f r a c t i o n a t i o n by gas-liquid •i^f- 71 "partition chromatography , following complete acetylation - 38 - of free primary hydroxyl groups, gave one zone on a column of 20$ S i l i c o n e G.E.S.F. 96 on f i r e b r i c k at 190° , which was shown to be a mixture of two components by t h i n layer chromato- graphy. Successful separation of the mixture of products from the oxo reaction was subsequently effected by deacety- l a t i o n , and chromatography of the r e s u l t i n g mixture of polyols. •J Following deacetylation by sodium methoxide i n metha- n o l ^ 2 , and removal of sodium ions with Amberlite IR-120 (H+) cation exchange r e s i n , a syrupy product was obtained, whose inf r a r e d spectrum confirmed the complete removal of 0-acetyl groups, and which showed a strong, broad hydroxyl band. Preliminary examination by descending paper chromatography revealed that the deacetylated product comprised mainly two components, detectable on spraying with perlodate- 73 >' Sc h i f f reagent ; With t h i s spray reagent, compounds having an oc-glycol group show as purple spots on a white background, by v i r t u e of t h e i r oxidation by periodate to a dialdehyde, which restores the colour of the S c h i f f re- agent. After development for about hO hours, using a s o l - vent system of water-saturated 1-butanol containing 5$ ethanol, the two major components of the mixture were suf- f i c i e n t l y f a r apart to enable t h e i r separation to be effected on a preparative scale. A preparative separation of the two components of the polyol mixture was c a r r i e d out by applying the material, i n methanol solution, to several - 39 - large sheets of Whatman's No. 1 paper prepared f o r descending chromatography. The progress of the separation was followed by detecting the positions of the zones on control chroma - tograms which were developed i n the same tank, and was allowed to continue u n t i l the two components of the mixture were near the bottom edge of the sheets, thereby achieving maximum separation. Three narrow test s t r i p s were cut from ft 73 each large sheet and sprayed with the zone-locating reagent i n order to ensure that the position of each zone was determined accurately; the material so l o s t represented from 5 to 10$ of each component. The zones thus located were exhaustively extracted with aqueous methanol; the two fractions thus Isolated, both i n i t i a l l y i n the form of syrups, were found to be free from contamination by the other on rechromatography of a portion on paper. Of the two chromatographically pure fractions r e- su l t i n g from the above separation, the faster moving com- ponent w i l l be designated Fraction I, and the slower, Fraction I I . (b) Characterisation of Fractions I and II From an amount of 400 mg of the mixture obtained on deacetylation of the oxo product, 150 mg of Fraction I and 180 mg of Fraction II were recovered from the chromatograms. Allowing for a loss of approximately 10$ of the material _ 40 - i n i t i a l l y applied (as a re s u l t of zone l o c a t i o n ) , the com- bined fractions therefore represented about 90$ of the mix- ture. B ( a ) n * ] 2 ° ( b ) Hp l JD m , p ' Fraction I 0.47 -13° 102° Fraction II 0 . 4 l -44° — (a) - i n water-saturated 1-butanol + 5$ ethanol, at room temperature (b) -ifo|water Fraction I, a f t e r c r y s t a l l i s a t i o n to constant melt- ing point, gave an elemental analysis corresponding to an empirical formula of CgH^O^. Acetylation of Fraction I with acetic anhydride-pyridine gave a syrup which could not be c r y s t a l l i s e d , although t h i n layer chromatography showed the homogeneity of the product. This f r a c t i o n was characterised as the p-nitrobenzoyl d e r i v a t i v e , m.p. 2 1 5 ° , |̂ <xjp - 5 0 ° . The elemental analysis of the c r y s t a l l i n e derivative corresponded to the replacement of three hydrogens of Fraction I by three p-nitrobenzoyl groups. Fraction II r e s i s t e d attempts at c r y s t a l l i s a t i o n , but r e a d i l y formed a c r y s t a l l i n e d e r i v a t i v e , m.p. 80-81° , {^oc]D -41°, on acetylation with acetic anhydride-pyridine. This derivative gave an analysis corresponding to C H-̂ gOy, - 41 - the t r i - O - a c e t y l derivative of Fraction I I . In order to obtain Fraction II i n a highly p u r i f i e d form, a portion of the c r y s t a l l i n e acetate was deacetylated by the action of methanolic sodium methoxide and the product was c a r e f u l l y i s o l a t e d i n the usual way to give a syrup which did not c r y s t a l l i s e , but which analysed s a t i s f a c t o r i l y f o r CgH 1 2 04« ( i i i ) I d e n t i f i c a t i o n of Fractions I and II (a) Periodate Consumption Information on the structure of a carbohydrate can 7 4 be gained from a study of i t s reaction with periodate ion' . The chief a n a l y t i c a l a p p l i c a t i o n of t h i s reaction i s i n the determination of the number of adjacent hydroxyl groups i n the molecule, as each ex.-glycol group consumes one molecular proportion of periodate, the r e s u l t i n g fragments being formaldehyde, formic acid or a substituted aldehyde according to the location of the p a r t i c u l a r group undergoing oxidation. CHOH CHGH I O 4 HCHO + HCOOH + CHO 36 R R Other groups such as oc-hydroxyaldehydes, oc-hydroxyketones and cx-amino-alcohols are also cleaved by periodate. - 42 - Various methods are i n common use f o r following the " ] ' 75 " reaction of v i c i n a l hydroxyl groups with periodate ion , usually involving t i t r i m e t r i c procedures and requiring the destruction of appreciable amounts of material. The spectro- 7 6 photometric method of Dixon and L i p k i n 1 , requiring only 8 6 10" to 10~ mole of sample, appeared to be eminently s u i t - able f o r determining the number of v i c i n a l hydroxyl groups i n Fractions I and I I , l i m i t e d amounts of which were a v a i l - able. These workers found that consumption of periodate may be followed spectrophotometrically at 223 ttyx, at which 77 wavelength i t s absorption i s at a maximum . No inaccuracies were introduced by the r e l a t i v e l y small absorption of iodate ion i n t h i s region, nor by absorption of unconjugated car- li bonyl compounds formed during the reaction. Following t h i s procedure, the decrease i n absorbance at 223 rn/A. of a solution of Fraction I (0.439 x 10 M on the basis of a molecular weight of 200) containing an excess -4 of sodium periodate (0.942 x 10 M) was measured over a period of several hours. Each reading (A) on the Beckman Model DU Spectrophotometer was accompanied by a reading 4 (B) of a control solution 0.942 x 10 M with respect to -4 periodate ion, and also of a solution containing 0.439 x 10 M of Fraction I (C). Thus (B+C)-A was a measure of the decrease i n absorbance due to consumption of periodate ion by Fraction C, I. Then (^+c),"A w a s the f r a c t i o n of the known amount of o - 43 - added periodate which was consumed by the carbohydrate at any time, where C Q was the measured absorbance of the perio- date solution at zero time (this changes slowly with time 77 x because of variations i n pH, temperature ). It was found that A reached a constant value equivalent to the consump- t i o n of 0.90.mole of periodate ion per mole of Fraction I. A s i m i l a r series of measurements with Fraction II gave a 1 • i value of 0.95/'moles of periodate per mole of substrate. With due allowance f o r the micro scale of these analyses, these r e s u l t s therefore showed the presence of two v i c i n a l hydroxyl groups i n each compound. (b) Structures of Fractions I and II Thus f f r , the Information obtained on Fractions I and II indicated that both contained three hydroxyl groups/ two of which were v l c l n a l l y situated. On t h i s evidence, and from a consideration of the l i k e l y mode of addition of carbon monoxide and hydrogen to 3 , 4-di-O-acetyl-D-xylal (22) i t could be assumed that the reaction had followed the ex- pected course, and a hydroxymethyl group had added to the double bond. On t h i s basis, four Isomers (44) to (47) were possible, a l l of which contain two v i c i n a l secondary - 44 - hydroxyl groups-in addition to one primary hydroxyl group, and therefore could not he distinguished on the avai l a b l e evidence. Structures (44) and (45), r e s u l t i n g from the addition of a hydroxymethyl group at C - l of the g l y c a l , are l,5-anhydro-4-deoxy-hexitols, d i f f e r i n g i n the configura- t i o n of carbon 5» i n accordance with accepted nomenclature^®, (44) and (45) are preferably drawn i n the following way Hydroxymethylation at C-2 of the g l y c a l would give either of the branched-chain isomers (46) and (47), which are l,5-anhydro-2-deoxy-2-hydroxymethyl-pentitols, d i f f e r i n g i n configuration of the side chain at C-2. I t was possible to d i s t i n g u i s h between the straight chain and branched chain structures merely from a considera- t i o n of the resonance positions and r e l a t i v e i n t e n s i t i e s of the proton signals observed i n the nuclear magnetic resonance spectra of Fractions I and I I , measured i n deuterium oxide so l u t i o n . In t h i s solvent, hydroxylic protons are r a p i d l y exchanged and are resolved into one sharp H-O-D s i g n a l ; the OH OH (44) (45) - 45 - spectrum i s thereby s i m p l i f i e d . Consideration of structures (44) to (47) shows that i n both straight chain and branched chain isomers are present (a) methylene protons o t t o an oxygen, C-CHg-O-, and (b) a methine proton OC to an oxygen function, ' In ̂ he - straight chain isomers (44) and (45), a methylene group i s also present which i s flanked by two carbon atoms ; (c) C-CH2-C, whereas i n the branched chain isomers (46) and (47) a t e r t i a r y hydrogen, rather C than a methylene group, i s present; (d) G-6H-C. I t i s fundamental to nuclear magnetic resonance spectroscopy that the resonance po s i t i o n , or chemical s h i f t , of a proton depends upon i t s precise chemical environment. Ample evidence i s a v a i l a b l e to indicate with certainty that protons of types (a) and (b) above w i l l resonate at a lower magnetic f i e l d than w i l l type (e) and (d) protons, because the inductive ef f e c t of the adjacent oxygen atom w i l l reduce the electron density around these protons. Jackman^^ tabulates t y p i c a l values f o r methylene and methine protons with an oc-oxygen function as l y i n g between 6 .15 and 6.6© T ( 8 = 3 . 8 5 - 3 . 4 0 ppm), whereas the corresponding protons i n saturated hydro- carbons resonate around 8 .5 t (1.5 pptn), with deshieldlng from -substituents r e s u l t i n g i n a comparatively minor s h i f t to lower f i e l d . Consequently, i n DgO solution, i t was anticipated that the non-hydroxylic hydrogens of the straight chain isomers (44) and (45) would exhibit two types of sig n a l s : a group - 46 - at lower f i e l d of r e l a t i v e i n t e n s i t y 7, corresponding to type (a) and (b) protons, and a higher f i e l d group of r e l a - t i v e i n t e n s i t y 2, corresponding to %f\e two methylene protons, (c ) . On the other hand, with the branched chain isomers J46 ) and (47) the lower field.group would have a r e l a t i v e i n t e n s i t y of 8 , whereas the single t e r t i a r y hydrogen (d) would resonate at higher f i e l d with an i n t e n s i t y of 1. The observed n.m.r. spectra of Fractions I and II both showed the anticipated separation of signals into lower and higher f i e l d groups, (in addition to the single H-O-D peak at 8 = #.73 ppm). In both cases the area enclosed by the lower f i e l d group of signals ( S = 2 .9 -4 . 2 .ppm) was 3.5 times that of the group at higher f i e l d ( S = 1 .1-2.2 ppm). This c l e a r l y demonstrated that both Fractions I and II were isomeric 1 ,5-anhydro-4- deoxy-hexitols (structures (44) and ( 45 ) ) . Additional i n - formation regarding the stereochemistry of these compounds can be derived from a consideration of the m u l t i p l i c i t i e s of the higher f i e l d signals? t h i s aspect i s discussed i n more d e t a i l l a t e r . Further confirmation that Fractions I and II were isomeric l ,5-anhydro -4-deoxy-hexitols was based on the follow- ing argument. Periodate cleavage of the oc - g l y c o l group . • i n a l l four structures (44) - (47) would furnish a dialdehyde which on reduction would give a structure having three p r i - mary hydroxyl groups and an ether linkage. However, whereas from (44) and (45) a pair o f enantiomeric t r i b l ethers (48) and (49) would b e obtained, e a c h having one centre of asym- metry (-fc), t h e same reactions applied to either branched chain isomers, (46) or . (47), would r e s u l t i n the same op- t i c a l l y - i n a c t i v e t r i o l e t h e r (50) being formed. onpn KC-O-CH^CI-ph! ChjpH (48) CrjpH HOChp-jjO-C-H .ChjpH (49) OH (45) (46) or (4?) 9 4P H CH-a-^-0-CH2CH2OH Chj,OH (50) Both Fractions I a n d I I w e r e separately treated, i n t h i s way, according to a p r o c e d u r e s i m i l a r to that described 80 by von Rudloff a n d co-workers . A sample of each polyol was oxidised w i t h a 5 0 $ e x c e s s o f periodic acid u n t i l the o p t i c a l rotations of t h e s o l u t i o n s were constant. After n e u t r a l i s i n g with b a r i u m carbonate the solutions were treated with an aqueous solution of s o d i u m borohydride; cations were removed with -Amberlite- IR-120(H"*) r e s i n , and borate ion was - 48 - v o l a t a l i s e d as the methyl ester was a syrup. 81 The product In both cases The t r i o l ether obtained from Fraction I had D - 1 9 ° , and that from Fraction II had [p^-p +17 0. Both products had i d e n t i c a l n.m.r. spectra, measured i n deuterium oxide solu- t i o n , with a group of signals 8 = 3.5 - 3*9 ppm, and a higher f i e l d group at 8 - 1.5 - 2 .0 ppm, the r e l a t i v e areas of the two groups being i n the r a t i o of 9 : 2 . Hence the products r e s u l t i n g from the periodate oxidation and sodium borohydride reduction of the two polyols were c l e a r l y the enantiomeric D- (48) and L- (49) forms of 2-deoxy -3 -0-(2-hydroxyethyl)- g l y c e r o - t e t r i t o l . I t was noted that, i n the course of an investigation of the products r e s u l t i n g from the hydrogenolysis of methyl Oc-D-glucopyranoside (51) under conditions of high tempera- ture and pressure, and i n the presence of a copper chromite c a t a l y s t , von Rudloff and co-workers® 0 had i s o l a t e d , i n addition to several other products, a 2-hydr6xymethyl-4 ,5-dihydroxy- tetrahydropyran (52) ; (considered as a carbohydrate, (52) i s a it • • 1,5 -a nhydro -4j|deoxy -hexi t ol) CH20H HO ,OH Ohl 6CH, (51) + other products 40 CKpH - 4 9 - The molecular structure of (52) (which l a t e r proved to be a mixture of stereoisomers ) was demonstrated by periodate oxidation to a dialdehyde which was then reduced with sodium borohydride to an open chain t r i o l ether ( 5 3 ) , characterised as the t r i a c e t a t e . The structure of the t r i o l ether was proved i n two ways: (a) ethyl iodoacetate was condensed with d i e t h y l L-twaiate i n the presence of sodium, and the r e s u l t i n g t r i e s t e r (54) was reduced with l i t h i u m aluminum hydride COOEt I ICH .COOEt -+ CH„ 2 |_ 2 - HO-C-H I COOEt Na COOEt I CH <3H I 2 LiAlHh CH 0 > I 2 CH-0-CH0.COOEt I 2 COOEt (54) CH-0-CHo.CH^OH I 2 2 CH20H (53) /x 83 (b) cleavage of the t r i o l ether with boron t r i c h l o r i d e i n acetic anhydride and subsequent deacetylation gave ethane- d i o l and 1 , 2 , 4-butanetriol CHo0H I 2 CH^ BC1. CH-0-CHp.CH OH I CH OH (53) CH^OH I 2 CHOH CH OH 2 H0CHiCHo0H 2 ^ 2 42 - 50 - The product (52) from the hydrogenolysis of (51) Op was further investigated by Gorin , who fractionated the material by c e l l u l o s e column chromatography Into two i s o - meric l , 5-anhydro-4-deoxy-hexitols. Periodate oxidation and borohydride reduction of both components afforded the same t r i o l ether, which demonstrated that the configuration of the hydroxymethyl group was the same i n each case, and therefore the two compounds d i f f e r e d i n the configuration of the v i c i n a l hydroxyls on the r i n g . The t r i o l ethers obtained by Gorin were levorotatory ( [°<GD - 2 0 ° , - l 6 ° ) , and were characterized as the tris-p-nitrobenzoates, melting points 103-104°, 98-102°, s p e c i f i c rotations - 2 6 ° , - 2 2 ° . The enantiomeric t r i o l ethers obtained from Fractions I and I I , which f o r the sake of convenience w i l l be designated as compounds III and IV respectively, were converted to p-nitrobenzoyl derivatives according to the procedure of ftp Gorin° . The levorotatory t r i o l ether I I I from Fraction I gave a c r y s t a l l i n e derivative m.p. 102-103° , 0*0 D - 2 8 ° , which did not depress the melting point of Gorin's t r i s - p - n i t r o - o j i benzoate , and the dextrorotatory t r i o l ether IV, from Fraction I I , furnished the enantiomeric deri v a t i v e , m.p. 102-103° , with an equal and opposite s p e c i f i c r o t a t i o n . As the previous ftO work of von Rudloff and co-workers had c l e a r l y demonstrated the positions of attachment of the hydroxymethyl group and the two secondary hydroxyl groups to the tetrahydropyran - 51 - r i n g of ( 5 2 ) , these re s u l t s provided addi t i o n a l and con- clu s i v e proof that Fractions I and II were isomeric 1 ,5- anhydro -4-deoxy-hexitols. One implication of the demonstrated r e l a t i o n s h i p between the two isomeric anhydrodeoxyhexitols derived from methyl cx-D rglucopyranoside (51) and Fraction I was that, assuming no inversion of configuration occurred at C-5 of • the glycopyranoside r i n g of (51) during the hydrogenolysis reaction, then Fraction I must have the D configuration at C-5 ( 4 4 ) , and consequently Fraction II must have the L- configuration at t h i s centre ( 4 5 ) . However, such an assump- t i o n was not considered s u f f i c i e n t l y j u s t i f i a b l e without further confirmation, f o r two reasons: 80 1. The hydrogenolysis reaction of von Rudloff and co-workers was c a r r i e d out under extremely vigorous conditions, and i n general i t has been found that these reactions are accompanied 8s 86 by considerable configurational changes 2. Assuming the secondary hydroxyl groups of Fractions I and II were unchanged i n configuration as a r e s u l t of the oxo reaction, then on the basis of the evidence described above, of the two possible l , 5-anhydro - 4-deoxy-hexitols, one of the two fractions must be 1,5-anhydro-4-deoxy-D-arablno- h e x i t o l ( 4 4 ) , and the other must be the corresponding L-xylb- Isomer (45). Both the anhydrodeoxyhexitols Isolated by i o r l n - 52 - HO- i -H .(J> H-C-OH HO- i -H (44) = H-C ^ ihiOH' (45) = Q were assumed to be of the D-series as they were derived from a D-glucopyranoside (51) , and on the basis of cer t a i n evidence, which w i l l be discussed i n more d e t a i l below, were assigned the D-arablno-(44) and D-lyxo-(6o) configurations. However, the properties of the isomer which was considered to have the D-arabino-configuration (44) did not resemble those of either Fraction I or Fraction I I . (iv) Configurations of Fractions I and II (a) Stereochemistry at C-5 The stereochemistry at C-5 of the 1 ,5-anhydro-4- deoxy-hexitols I and II was established with certainty once i t was known which of the enantiomeric t r i o l ethers III and IV was 2-deoxy - 3 - 0 - ( 2-hydroxyethyl)-D-glycero-tetritol (48), and which was the L-isomer (49). Attempts were made to obtain one of the optically-pure enantiomers by an unequivocal route J t h i s was r e a l i s e d by the preparation of the L-isomer (49) from a structure of known stereochemistry, 1,4-anhydro- 5-deoxy-B-arabino-hexitol ( 5 9 ) , obtained from the known com- pound 3,6-anhydro-2-deoxy-B-lyxo-hexose (58) on reduction with sodium borjfehydride. The anhydrodeoxy-sugar (58) was obtained by the procedure of Poster and co-workers0"^ from methyl 2-deoxy-oc-B-galactopyranoside ( 5 5 ) , which i n turn . was prepared from B-galactal ( 1 9 ) . HO CH2OH OH (19) (49) HO CHpH OH (55) CH^OH OMe GHO CHpTs HO J Ov (59) (58) OMe (56) OMe (57) B-galactal (19) was obtained from 3,4,6-tri-O-acetyl- B-galactal (4G) on deacetylatlon with methanolic sodium metho- 88 xide J t r a n s e s t e r i f i c a t i o n of (40)(and other acetylated 89,66N g l y c a l s ) under these conditions i s unusually slow, taking 2 days f o r completion. B-galactal c r y s t a l l i s e d r e a d i l y from - 54 - the residue remaining a f t e r n e u t r a l i s a t i o n and evaporation of solvent, upon extraction with ethyl acetate. Preparations of methyl 2-debxy-D-galaetopyranoside (55) have been reported by Tamm and Reichstein and by Overend and co-workers ; i n each case D-galactal (l4) was f i r s t converted to 2-deoxy-D-galactose by the addition of water across the double bond i n d i l u t e sulphuric acid solution, and the sugar was then methylated i n the presence of hydrogen chloride, the more stable cx-glycoside (55) being i s o l a t e d i n c r y s t a l l i n e form. The l a t t e r workers also observed the d i r e c t formation of (55) when D-galactal was treated with 3 .3$ methanollc hydrogen chloride'} 4and 91 Foster, Overend and Stacey obtained the oc-methyl glyco- side on a q u a l i t a t i v e scale when a solution of D-galactal i n 0.4$ methanollc hydrogen chloride was allowed to reach r o t a t i o n a l equilibrium over 43 minutes. The one step con- version of D-galactal to methyl 2-deoxy-oc-D-galactopyranoslde was used on a preparative scale i n t h i s work. The change i n o p t i c a l rotation of a solution of D-galactal (19) i n methanol containing 0 .3$ hydrogen chloride was followed, and found to be constant a f t e r about 50 minutes. A syrup was i s o l a t e d from t h i s reaction which c r y s t a l l i s e d r e a d i l y on adding a small volume of acetone; the product was i d e n t i - f i e d as (55) from i t s melting point of 112-114°, (Overend and co-workers^ 2 found 112-113°)• - 55 - The preparation of 3>6-anhydro-2-deoxy-D-lyxo-hexose (58) from (55) was car r i e d out by the method of Foster and co-workers^ 7; these workers applied the known procedure of Haworth and co-workers^ 2 for the preparation of 3,6-anhydro- sugars, involving a l k a l i n e treatment of the 6-Q-p-tolysulphonyl derivatives of the methyl glycosides, with formation of the corresponding methyl 3 ,.6-a nhydro-glycosides. Thus reaction of (55) with 1 molar equivalent of p-toluenesulphonyl chloride under conditions favourable f o r unimolar sulphonylation 0^ gave the syrupy 6 - 0-p-tolylsulphonyl derivative ( 5 6 ) , which on treatment with base was converted to methyl 3 ,6-anhydro-2- deoxy- cx-D-galactopyranoside (57)* m.p. 7 6 - 7 7 ° , 2 5 + 9 4 0 (Foster and co-workers^ 7 give m.p. 80°, [pc]15 ^ 9 8 ° ) . As proof of the configuration of the t r i o l ether (49) ultimately obtained from l ,4-anhydro -5-deoxy-D-arabino-hexitol (59) e s s e n t i a l l y depends upon a knowledge of the configuration at C-3 the anhydro-deoxy-galactoside ( 5 7 ) , i t i s necessary to consider the evidence f o r the structure of t h i s compound: (a) (57) was characterised as the c r y s t a l l i n e 4-0-p- toljsulphonyl d e r i v a t i v e ^ 7 . (b) When the syrupy 6 - 0-p-tolylsulphonyl derivative ( 5 6 ) , from which (57) was prepared, was heated with sodium iodide i n acetone solution, an equivalent amount of sodium p- toluenesulphonate was obtained, evidence that the tosyloxy group of (56) i s at the primary (C-6) p o s i t i o n ^ . (c) The mechanism of formation of anhydro-sugars i n basic medium i s well established, and has been reviewed by ?eaty . 3,6-Anhydro-sugars are of the hydrofuran'ol type, In which the anhydro-ring i s 5-membered; both t h i s type and the well known ethylene oxide (3rmembered) anhydrides are often prepared by treatment of a t o s y l derivative with base. The reaction involves an intramolecular exchange of anions at a carbonium ion, whereby a tosyloxy group i s replaced by anionic oxygen, and i s most c l e a r l y i l l u s t r a t e d with r e f e r - ence to the ethylene oxide type of anhydride. Thus anhydride 44 formation involves inversion of configuration at the carbon atdm bearing the poten t i a l leaving group (which may also be halogen or other mineral acid ester group), but no inver- sion of configuration occurs at the hydroxylic carbon. It follows that the displacing ion must have a trans arrange- ment with respect to the leaving group before an anhydride r i n g can form. The same p r i n c i p l e holds i n the formation of 3 ,6-anhydrides as a re s u l t of replacement of a tosyloxy group at C-6 by nucleophilic oxygen at C - 3 . As C-6 Is not asymmetric no inversion of configuration i s apparent; however, . - 57 - the formation s t i l l requires a trans d i s p o s i t i o n of the entering and leaving groups. For s t e r i c reasons the free hydroxyl group at C-3 must be on the same side of the sugar rin g as the side chain, C - 6 . C l e a r l y , therefore, formation of >3,6-anhydro-sugars under these conditions r e s u l t s i n no c o n f i g u r a t i o n 1 inversion at C - 3 , and ( 5 9 ) , obtained from (58) on sodium borohydride reduction, must have the re- configuration at C-4. 3,6-Anhydro-2-deoxy-D-lyxo-hexose (58). was obtained, on treatment of the anhydrodeoxygalactoside (57) with d i l u t e acid at room temperature, as a syrup, t?<3D + 2 5 0 (reported r I o 8 7 |OcJ D+24 u ). Fost er and co-workers note that the anhydro- sugar (58) exists i n the aldehydo-form, r e a d i l y restoring the colour to S c h i f f reagent, and a carbonyl band at about 1715 cm"1 was observed i n the infrared spectrum of the product. In t h i s respect (58) d i f f e r s from the corresponding derivatives of D-glucose and D-mannose, which exist i n thje furanose form, an arrangement of two els-fused 5-membered . - 58 - rings being i n a less-strained state than a 3 ,6-anhydride bridge across a pyranose r i n g . Such an arrangement i s not possible i n 3,6-anhydro-D-galactose and Its 2-deoxy-derivative (58), because of the orientation of the hydroxyl group at C - 4 . Reduction of (58) with sodium borohydride gave 1 ,4- anhydro-5-deoxy-D-arablno-hexitol ( 5 9 ) , as evidenced by the, disappearance of carbonyl. absorption i n the i n f r a r e d . The anhydrodeoxyhexitol ( 5 9 ) , which does not appear to have been 25 reported previously, had [Py £ + 2 1 ° , and was a syrup. It was characterised as the trls-p-nltrobenzoyl d e r i v a t i v e , m.p. 1 5 9 - l 6 0 O , CP'GQ2 - 9 6 ° . (59) was oxidised with excess G.1M periodic acid, under conditions si m i l a r to those employed for the cleavage of Fractions I and I I . By comparison with these previous reactions, the oxidation of (59) was slow,' as would be anticipated f o r a'trans ( X - g l y c o l group-on a •74 ~ 5-membered rin g . Sodium borohydride reduction of the re- s u l t i n g .dialdehyde then gave 2-deoxy -3-0~(2-hydroxyethyl)- L - g l y c e r o - t e t r i t o l (49) ; t h i s had M D + 1 7 ° , , and formed a tris-p-nitrobenzoate, m.p. 1 0 1 - 1 0 2 ° , C^clp+ 2 7 ° , whose melting point was., undepressed on admixture with the p- nitrobenzoate of the t r i o l e t h e r IV obtained from Fraction I I . Thus, Fraction II must have the L-configuration at C - 5 , and Fraction I, which on cleavage and reduction afforded the enantiomeric t r i o l ether I I I , must have the.©-configuration 82 at C - 5 . These res u l t s confirmed Gorin's assumption that no inversion occurred at.C-5 of methyl Oc-D-glucopyranoside (51) during the course of the hydrogenolysis of t h i s compound® 0, and therefore both polyols i s o l a t e d from t h i s reaction were of the D-series. (b) Configurations of Secondary Hydroxyls of Fractions I and II An element of disagreement with Gorin's results s t i l l prevented a completely unequivocal assignment of the D- arablno - ( 4 4 ) and L-xylo - ( 4 5 ) configurations to the two an- hydrodeoxyhexitols, Fractions I and II respectively.. Gorin® 2 found that his two polyols, which f o r the sake of convenience w i l l be designated as X and Y, consumed lead tetraacetate at d i s t i n c t l y d i f f e r e n t rates, X being oxidised four times fas t e r than Y i n the i n i t i a l stages. As both polyols were otherwise s t r u c t u r a l l y s i m i l a r , t h i s was taken as e v i d e n c e ^ that X contained a c i s , and Y a trans o c - g l y c o l group. Assuming that the configuration at C-5 was D i n both cases (an assumption j u s t i f i e d by our r e s u l t s ) , a t o t a l of four possible isomers, two c i s and two trans, was then considered. D-lyxo D-ribo D-arabino D-xylo (62) (60) (61) (44) - 60 For polyols X and Y, Gorln found s p e c i f i c rotations of -50° and 4 1 9 ° respectively. These values were compared with those calculated by application of the p r i n c i p l e of o p t i c a l superposition, f i r s t postulated by van't Hoff and l a t e r applied to carbohydrates by Hudson 0^ i n the form of his Isorotation Rules. For each of the four possible isomeric anhydrodeoxyhexitols ( 4 4 ) , ( 6 0 ) , ( 6 l ) and ( 6 2 ) , a value f o r the molecular rotation was derived by reference to a pair of structures-of known molecular rotations, whose i n d i v i d u a l asymmetric centres "cancelled out" except f o r those which were also present i n the anhydrodeoxyhexitol. This i s best i l l u s t r a t e d by two examples, selected from the six tabulated by Gorln, which pertain to the D-lyxo-(6o) and D-arabino - (44) isomers. (-b) HO~(>H (-c) HO-(U-t (-e) H-C- ' 0. CH^OH (60) (+a) ^ (-b) H O - C - H (-cj H O - C - H (-d) HO - £ - -H (-e) H-Q- O CH 2 OH (63) (-a) H 3 C ° y ^ (-b) H O - C - H ^ (-c) H O - C - H M ) (-e) H,OH (64) Let the three i n d i v i d u a l centres of asymmetry i n 1,5-anhydro-4-deoxy-D-lyxo-hex!t o l (60)$ at carbon's 2 , 3 and S-61 - 5, "be assigned i n d i v i d u a l contributions to the t o t a l mole- cular r o t a t i o n of -b, -c and -e. The corresponding c o n t r i - butions of the f i v e asymmetric centres of methyl cx-D- tallopyranoside ( 6 3 ) , whose molecular ro t a t i o n , [M] T , i s +4,070°, are +a, -b, -c, -d and -e, and i n methyl £>-D- mannopyranoside (64), ([M] M= - 1 3 , 3 9 0 ° ) , are -a, -b, -c, +d and -e. Therefore [M] T +[H) M = 2(-b-c-e), so that [M]^ , the calculated molecular rotation of the D-lyxo-lsomer, i s [M]T+ Mm/2 = - 4 , 6 6 0 ° , whence the calculated s p e c i f i c rota- t i o n , M/to, i s - 3 1 ° . In a si m i l a r manner the molecular rotation [M] of the D-arabino-isomer (44) was obtained by halving the sum of the molecular rotations of methyl o^-D-altropyranoside (65) ( M A = 24,420°) and of methyl (3 -D-idopyranoside (66) ( [MJJ = - 1 8 , 4 3 0 ° ) . This gave a value of +20° f o r • the s p e c i f i c r o t a t i o n of the D.-arabino-isomer (44) . (-b) HO-C-H (+c) H-C-OH (-e) H - C ^ ^ ChyDH (44) (+a) (-b) (+c) ( +d) (-e) H V / 0 C H 3 H O ^ - H H-C-OH H-C-OH H-Q- 6 (65) HOH 2 (-a) (-b) (+c) (-d) C-e) H3CO /H T HO-C-H H-C-OH HO-C-H 6 H O H "•2.. (66) - 62 - Calculated values f o r the D-rlbo - ( 6 l ) and B-xylo - (62) Isomers, from the known molecular rotations of various other methyl hexopyranosides and 1 , 5-anhydrohexitols, were as tabulated below. Observed [p3 D Calculated [c*ID of isomeric 1, 4-deoxy-D-hexitols , 5-anhydro- X Y D-ribo ( 6 l ) D-lyxo(6o) D-xylo(62) D-arabino(44) , - 5 0 ° o 19 ~ ' 46° " - 3 1 ° 106° -34° 20° 67° Comparisons of the observed rotations of polyols X and Y with those calculated f o r the four possible stereoisomers then led to the conclusion that polyol X was probably 1 ,5- anhydro-4-deoxy-D-lyxo-hexitol (60), and that Y was probably the D-arabino-isomer (44) . This conclusion therefore d i s - agreed with the assignment of the D-arabino- configuration to Fraction I, which had a s p e c i f i c r otation of - 1 3 ° . Furthermore, Gorin's polyol Y was characterised as the tris-p-nltrobenzoate, m.p. 1 1 5 - 1 1 9 ° , 5 9 ° , whereas Fraction I gave a p-nitrobenzoyl derivative with m.p. 215° and M D - 5 0 ° . It was then necessary to consider the p o s s i b i l i t y that configurational lnversiomeihad occurred during the course of the oxo reaction of 3 , 4-di-O-acetyl-D-xylal. Certain circumstancial evidence can be mentioned which indicated that the D-threo-configuration of the six-membered ring was retained during t h i s reaction. Thus when 3 , 4-di-O-acetyl- D-xylal was reacted with only 2 moles of synthesis gas, a quantity of unchanged g l y c a l was i s o l a t e d . The trans-arrangement of the cx - g l y c o l system i n Fractions I and II was indicated by the fact that neither compound formed a derivative with acetone, as i t i s known that a c i s configuration i s a pre- r e q u i s i t e of isopropylidene acetal formation. Some exploratory experiments were carr i e d out with a view to synthesising the D-arabino-(44) and L-xylo-(45) isomers of l,5-anhydro-4-deoxy-hexitol by an a l t e r n a t i v e route from 3,4-di-O-acetyl-D-xylal, whereby the p o s s i b i l i t y of r i n g inversions was absent. The recently reported pre- 98 ^ paration by Goxon and Fletcher of. a 2,6-anhydro-heptItol by reduction, followed by deamination of 2 , 3,4 , 6-tetra-0- acetyl-£ -D-glucopyranosyl cyanide suggested the p o s s i b i l i t y of employing a s i m i l a r sequence bf reactions subsequent td the introduction of a cyano- group at C - l of the g l y c a l . I t i s claimed^ 0 , that hydrogen cyanide reacts with 2-alkoxy derivatives of 2 ,3-dihydro-4H-pyran ( l ) , i n the presence of a basic c a t a l y s t , with formation of 6-cyano-2-alkoxy- tetrahydropyrans. However no reaction was observed over several hours when 3,4-di-O-acetyl-D-xylal (22) was dissolved i n anhydrous hydrogen c y a n i d e 1 ^ i n the presence of sodium cyanide; apparently the double bond of glycals i s i n s u f f i - 101 c i e n t l y activated f o r t h i s addition to take place . It was noted that 2-cyano-tetrahydropyran has been prepared - 64 - by the addition of hydrogen chloride to the double bond of 2,3-dihydro-4H-pyran ( l ) , followed by replacement of chloride .102 by cyanide on r e f l u x i n g with s i l v e r cyanide i n ether J lithium aluminum hydride reduction of the cyano- derivative gave 2-aminomethyl-tetrahydropyran. A s i m i l a r preparation of the cyano- derivative from 2-bromo-tetrahydropyran has been effected by the action of s i l v e r or mercuric c y a n i d e 1 0 ^ . When 3 , 4-di-O-acetyl-D-xylal (22) i n cold benzene solution was saturated with dry hydrogen chloride a syrupy product was obtained on removal of solvent which exhibited no absorption at 1640 cm - 1, t h i s i n d i c a t i n g that quantitative addition of HC1 had occurred across the double bond of the g l y c a l . Reaction of the product with mercuric'cyanide i n nitromethane 0^ gave a dark coloured product which could not be p u r i f i e d by chromatography. Although no absorption was present i n the 2000-2300 cm"1 region of the i n f r a r e d spectrum , analysis showed the presence of 2.7$ nitrogen i n the chromatographed product. Refluxing the H C 1 - addition 102 product with s i l v e r cyanide i n ether apparently resulted i n no reaction other than p a r t i a l regeneration of ( 2 2 ) . Though these experiments were unsuccessful, they de- monstrated the fact that hydrogen chloride adds r e a d i l y across the double bond of 3 , 4-di-O-acetyl-D-xylal ( 2 2 ) , with formation of 3 , 4-di - 0-acetyl - 2-deoxy-D-xylopyranosyl chloride ( 6 7 ) . D a v o l l 1 0 ^ has reported the s i m i l a r addition of HC1 o (22) HC1 H,CI > 46 AcO (67) (and HBr) to I f , 4 - d i - 0 - a c e t y l-D~ a r a b i n a l ( 3 2 ) ; the product was not stable, and was not characterised. It was observed that the addition product of hydrogen chloride to (22) also tended to decompose on standing at room temperature (in chloroform so l u t i o n ) , and absorption at 1640 cm"^ reappeared i n the i n f r a r e d spectrum. In t h i s respect the glycals resemble 2 ,3-dihydro-4H~pyran (l) which q u a n t i t a t i v e l y adds hydrogen chloride and hydrogen bromide, but the r e a d i l y dehydrohalogenated products have not been i s o l a t e d , and are 106,107 normally reacted i n s i t u . An a l t e r n a t i v e approach to the problem i n hand was based on this observed addition: a methyl, rather than a hydroxymethyl group, was introduced at C - l of the six -member ed r i n g of 3 , 4-di - 0-acetyl-I)-xylal ( 2 2 ) , and the products thereby obtained were i d e n t i f i e d with those r e s u l t i n g from reduction of the terminal hydroxy- methyl groups of the two l ,5-anhydro-4-deoxy-hexito.ls, Fractions I and I I . The action of a l i p h a t i c or aromatic Grignard reagents on s p e c i f i c functional groups of carbohydrates has been developed by Bonner and co-workers as a v e r s a t i l e t o o l - 66 - for the introduction of a l k y l or a r y l groups into various positions of the molecule. In general, two types of reaction are observed; 1. Normal Grignard addition involving carbonyl functions, such as lactones, esters and aldehydes, 2. Metathetical reactions. Of the second type, reactions of p a r t i c u l a r interest are those of Grignard reagents with polyacetyl-glycosyl halides. S t r u c t u r a l l y these compounds are herniacetal halides and therefore resemble oc -chloro-ethers, which have long been known to react with Grignard reagents i n the following 109 manner CI R< R-CH + R'MgX > R-CH + MgXCl 47_ OR OR Although the reactions of polyacetyl-glucosyl halides with Grignard reagents has been a subject of interest since 1 9 0 6 1 1 0 ' 1 1 1 ' 1 1 2 , i t was not u n t i l 19^5 that Hurd and Bonner 1 1 3 demonstrated metathesis involving the hemiacetal halide, according to equation 47_. The observations of previous workers 111 of addition product formation , or reaction only of the 110 a c e t y l groups with Grignard reagent was accounted for by the fact that an i n s u f f i c i e n t amount of the reagent was used, or that the reaction had been attempted at too low a temperature, - 6? - 114 Bonner showed that p r e f e r e n t i a l addition of Grignard r e- agent occurs at the ester groups before the metathetical reaction takes place to any appreciable extent; as each ester function takes up 2 moles of Grignard reagent (equation 48), complete reaction of a tetra-O-acetyl-glycosyl halide (equation 49) therefore requires the presence of nine moles RMgX^ of the reagent. Thus when 2 , 3 , 4 , 6-tetra-O-acetyl-cx -D- glucopyranosyl chloride (68) was refluxed i n ether with 12 moles of phenyl magnesium bromide, c r y s t a l l i n e ( 2 , 3 / 4 , 6 - tetra-O-acetyl- (3-D-glucopyranosyl)-benzene (69), accom- panied by the syrupy oc-anomer, was is o l a t e d a f t e r reacetyla. 113 tion" CHQAc OAc ;_+ 9C6H5MgBr — > CI AcO OAc C H OAc OAc / i-f ex -anomer ^ OAc (68) (69) Retention of configuration during the formation and subse- quent decomposition of the Grignard adduct to the ester function (equation 48) formed the basis of Bonner's i n v e s t i - gation of t h i s reaction, which was undertaken to determine whether the formation of the same type of product by the aluminium chloride catalysed glycosylation of aromatic 115 compounds was accompanied by intramolecular isomerisa- tions or inversions. That configuration i s retained on de- acetylation by Grignard reagents i s also a fundamental re- quirement of our experiment. This point i s proved experi- mentally by the fact that D-glucose i s obtained when 1 , 2 , 3 , 4 , 6 - penta-0-acetyl-6 -D-glucose i s reacted with Grignard re- 112 ~ agents . ' The metathetical reaction of Grignard reagents with the acetylated halldes of D-glucose, D-xylose and lactose was found by Hurd and Bonner to be an excellent general procedure for the preparation of aldopyranosyl derivatives of aromatic (benzene, toluene, naphthalene) and a l i p h a t i c (butane, isopropane) hydrocarbons. By the same procedure, Yoshimura and co-workers have prepared the analogous phenyl, benzyl, methyl, ethyl, propyl and butyl derivatives of 2- 116 amino-2-deoxy-D-glucose . The reaction, described below, of a Grignard reagent with a 2-deoxy-glyeosyl halide, derived from a g l y c a l , i s a hitherto unexplored aspect of t h i s general synthesis. 69 - Hydrogen chloride was added across the double bond of 3,4-di-O-acetyl-D-xylal (22) (equation 46), and an ethereal solution of the syrupy product (67) was added to a previously prepared solution of methyl magnesium bromide containing a two-fold excess of the Grignard reagent (approximately 10 moles). Reaction and decomposition of the reaction mixture was ca r r i e d out according to the procedure described by 108 Bonner , except that the product was i s o l a t e d at t h i s stage i n the deacetylated form, as i t was hoped to e f f e c t a separation by paper chromatography. However, chroma togiaris of the syrup which was i s o l a t e d , on developing with a variety of d i f f e r e n t sovent systems, revealed only one compact zone •73 on spraying with sodium periodate - S c h i f f reagent . Separation of t h i s component from any Impurities not revealed by the spray reagent was effected by preparative paper chromatography, and the p u r i f i e d material was then subjected to n.m.r. analysis. The spectrum obtained i n deuterium i oxide solution c l e a r l y showed that the apparently homogeneous product was a mixture of two compounds, both containing a C-CHg grouping, (and therefore indicated that the attempted reaction (equation 50) had succeeded). A \C A c O M OAc (67) 5CH^MgBr >HCI 1 ^ 50 - 70 - This was apparent from the presence of a pair of overlapping doublets (J= 6 c/s) at high f i e l d , 8 = 1.14 and 1.17 ppm. In addition to th i s absorption i n the C-methyl region of the spectrum, a multiplet around 8 = 1 .8 ppm was assignable to C-GH^-C, and a complex multiplet between 3 . 0 and 4 . 2 ppm c to the remaining hydrogens of the rings J the entire spectrum was thus i n d i c a t i v e of a mixture of the two anomeric forms of (2-deoxy-D-xylbpyranosyl)-methane. Retaining the previous nomenclature, these were l ,5-anhydro-4 ,6-dideoxy-D-arabino- h e x i t o l (70'), and 1,5-anhydro-4,6-dideoxy-L-xylo-hexitol (7l) • The r e l a t i v e Intensities of the two methyl doublets showed that the two components were present i n a r a t i o of about 3 : 2 . Resolution of t h i s mixture presented a problem which was solved by g a s - l i q u i d p a r t i t i o n chromatography (GLPC) of the acetylated anhydrodldeoxyhexitols. Acetylation of a portion of the mixture with acetic anhydride-pyridine gave a syrupy product which on t h i n layer chromatography showed the presence of two barely-resolved components! t h i s was further v e r i f i c a t i o n that the I n i t i a l product i s o l a t e d from the Grignard reaction was a mixture despite i t s apparent homogenlety on paper chromatography. After preliminary ex- periments to est a b l i s h suitable conditions, i t was possible to resolve the mixture of anhydrodideoxyhexitol acetates into two d i s t i n c t zones by GLPG, using a column (10' x W') - 71 - of 20$ Silcone GE-SF-96 on f i r e b r i c k at 180° , I f the amount of mixture applied to the column In solution was s u f f i c i e n t l y small. Resolution was less s a t i s f a c t o r y when a preparative separation was attempted (approximately 5 tng per i n j e c t i o n ) , and i t was necessary to c o l l e c t f r a c t i o n s enriched i n each component and rechromatograph these i n order to obtain each component i n a pure state, uncontaminated by the other Isomer. Each pure component gave an elemental analysis i n agreement with an empirical formula of C 1 0H 1gO^. The f a s t e r - moving the two components ( r e l a t i v e retention time 1 . 0 0 ) , - 2 1 ° , was more dextrorotatory than the slower component (r e l a t i v e retention time 1 . 1 0 ) , which had a s p e c i f i c r o t a t i o n of - 8 2 ° . Groups of signals present In the n.m.r. spectrum of each f r a c t i o n , measured In carbon tetrachloride solution, were r e a d i l y assigned from t h e i r resonance positions and r e l a t i v e i n t e n s i t i e s (in parentheses) to C-CH3 ( 3 ) , C-CHg-C ( 2 ) , acetyl (6) and ri n g hydrogens with an Oc-oxygen ( 5 ) , and confirmed the i d e n t i t i e s of the two components as i s o - meric 2 , 3-di - 0-a c etyl - 1 , 5-anhydro - 4 , 6-dideoxy-hexitols. On the basis that no inversion of configuration was possible as a r e s u l t of the reaction of (67) with the Grignard re- agent, one of the separated Isomers had the D-arabino- ( 7 2 ) , and the other the L-xylo- (73) configuration. - 72 - Ac o 0/pyridine (70) +• (71) > 51 These same two anhydrodideoxyhexitol diacetates were then synthesised by an alt e r n a t i v e route from the mixture of two isomeric 2 , 3-di - 0-acetyl-l , 5-anhydro - 4-deoxy-hexitols (74) and (75)* which formed the major part of the reaction product from the hydroxymethylation of 3 , 4-di-O-acetyl-D- x y l a l ( 2 2 ) , and from which Fractions I and II were obtained on deacetylation. This synthesis involved only the free primary hydroxyl groups at C - 6 , which were reduced to :methyl groups by a series of reactions which have been applied previously to the preparation of 6-deoxy-sugars, such as 6-deoxy-D-glucose 1 1 7 and 6-deoxy-D-galactose 1 1®, from the parent aldoses. CH 20H (22) CO/Hg^ QHpTs — Q C K O H AcQ OAc (75) - 73 - Nal (78) (79) Ra Ni OH ^ ((70) + ( 7 D ) 1 . 2. GLPC V (72) + (73) A portion of the syrupy mixture r e s u l t i n g from the oxo reaction of 3 , 4-di-O-acetyl-D-xylal (22) was treated with an excess of p-toleunesulphonyl chloride i n pyridine under standard conditions, and on working up the reaction mixture a product was obtained which, from an examination of the i n t e n s i t i e s of the c h a r a c t e r i s t i c resonances i n the n.m.r. spectrum associated with the p-tolylsulphonyl group, r e l a t i v e to those due to absorption by the acetyl groups, contained approximately 75$ of the isomeric 6-0-p- tolysulphonyl derivatives (76) and (77)• The crude product, i n acetone solution, was then heated i n a sealed tube i n the presence of sodium iodide, whereby the tosyloxy groups 1 were replaced by iodide to give the corresponding mixture of 6-deoxy - 6-iodo-derivatives (78) and ( 7 9 ) , and sodium p-toluenesulphonate was. pre c i p i t a t e d . On cooling and f i l - - 9 t r a t l o n , the amount, of sodium p-toluenesulphonate'''<Lsolated agreed c l o s e l y with the o r i g i n a l estimation, from n.m.r. - 74 - data, of the content of isomeric 6 - 0-p-tolylsulphonyl derivatives (76) and (77) i n the crude mixture. I s o l a t i o n of the reaction products (78) and (79) from residual sodium iodide was achieved by evaporation of the f i l t r a t e to dry- ness and extraction with ether. When the r e s u l t i n g product, i n s l i g h t l y basic methanol solution, wa3 hydrogenated at atmospheric pressure and temperature i n the presence of Raney n i c k e l , according to the procedure of Freudenberg 118 and Raschig , hydrogen was rapid l y absorbed and the mix- ture was simultaneously deacetylated. From t h i s reaction a syrup was is o l a t e d which showed only one zone by paper chromatography, i d e n t i c a l i n Rp value to that of the product obtained from the reaction of 3 , 4-di - 0-acetyl - 2-deoxy-D- xylopyranosyl chloride (67) with methyl magnesium bromide, and which, a f t e r p u r i f i c a t i o n by paper chromatography, had a n.m.r. spectrum which was e s s e n t i a l l y I d e n t i c a l to that previously obtained. That t h i s was a mixture of the same two D-arablno- (70) and L-xylo- (71) forms of 1,5-anhydro- 4 , 6-dideoxyhexItol was demonstrated by acetylation of the p u r i f i e d product and f r a c t i o n a t i o n by GLPC into two compon- ents. D -23° a n d - 8 1 ° , which had n.m.r. spectra i d e n t i c a l with those of the two Isomeric 2 , 3 - d i - 0 - a c e t y l - l , 5 - anhydro - 4 , 6-dideoxy-hexitols (72) and ( 7 3 ) , previously de- 1 • • • scribed. - 75 - Thus, the fact that the same two compounds were ob- tained by either route (equations 50 and 52) conclusively established that the secondary hydroxyl groups of Fractions I and II had retained the D-threo- configuration present i n 3 j 4-di - 0-acetyl-D-xylal. It was then possible to state with certainty that Fraction I was l ,5-anhydro-4-deoxy-I)- arabino-hexltol ( 4 4 ) , and that Fraction II was 1 ,5-anhydro-4- deoxy-L-xylo-hexitol ( 4 5 ) . Though not an essential part of t h i s proof, i t was of interest to i d e n t i f y i n d i v i d u a l l y the two isomeric d i - acetates (72) and (73) which were separated by GLPC. . I t i s known that primary hydroxyl groups react with p-toluene- sulphonyl chloride at a much f a s t e r rate than do secondary 119 hydroxyl groups ; i t i s therefore possible to p r e f e r e n t i a l l y sulphonylate a reactive primary position even when other vacant hydroxyls are present. The procedure of unimolar sulphonylation,.employing an amount of the sulphonyl chloride i n s l i g h t excess of that required for e s t e r i f i c a t l o n of the one reactive s i t e , was f i r s t introduced by Ohle and Dickhauser 93 and l a t e r improved by Levene and Raymond . Generally the reaction rate i s lowered by cooling, thereby further reducing the p o s s i b i l i t y of acylation of the secondary hydroxyls. 1,5-Anhydro-4-deoxy-L-xylo-hexito1 ( 4 5 ) , (Fraction II from the chromatographic separation of the deacetylated - 76 " hydroxymethylation product from 3 , 4-di-O-acetyl-D-xylal) was subjected to unlmolar t o s y l a t i o n ( l . l molar equivalents of p-toluenesulphonyl c h l o r i d e ) , with cooling i n ice J a f t e r 18 hours acetic anhydride was then added i n order to acetylate the secondary hydroxyl groups and f a c i l i t a t e i s o l a t i o n of the product. The r e s u l t i n g syrup, which showed absorption i n the i n f r a r e d spectrum (Characteristic of both a c e t y l and p-tolylsulphonyl groups (S= 0 stretching and aromatic C=C stretching v i b r a t i o n s ) , but not of hydroxyl, was not further p u r i f i e d . I t contained approximately 80$ of 2 , 3-di-O-aeetyl- l , 5-anhydro - 4-deoxy - 6 - 0 -(p-tolylsulphonyl)-L-xylo-hexitol (77)> as judged by the amount of sodium p-toluenesulphonate which was formed when the crude derivative was subsequently heated i n a sealed tube with sodium iodide i n acetone, during the conversion of (77) to the 6-deoxy - 6-iodo-derivative (79)• Reductive dehalogenation of (79) In the presence of Raney 121 n i c k e l , and simultaneous deacetylation, then gave 1 ,5- anhydro-4,6-dideoxy-L-xylo~hexito.l ( 7 3 ) . Following reacety- l a t i o n , t h i s was r e a d i l y i d e n t i f i e d by GLPC with'the f a s t e r - moving of the two components present i n the mixture of anhydro- dideoxyhexitol diacetates obtained by the two a l t e r n a t i v e routes described previously. Thus the more dextrorotatory of the two ( M D -21°) was 2 , 3-di - 0-acetyl - 1 , 5-anhydro - 4 , 6 - dideoxy-L-xylo-hexitol ( 7 3 ) , and the isomer having .[p3 D -82° was the corresponding D-arabino- isomer ( 7 2 ) . _ 7 7 j - (v) I d e n t i t i e s of Polyols X and Y 8 2 Prom the results discussed i n the previous pages, i t 82 was c l e a r that Gorin's assignment of the D-arabino- con- f i g u r a t i o n to the dextrorotatory l , 5-anhydro - 4-deoxyhexitol, (referred to previously as Polyol Y), Isolated from the products of hydrogenolysis of methyl oc-D-glucopyranoside 80 — (51) » was inco r r e c t . Prom a consideration of the available evidence, i t was possible to deduce the i d e n t i t y of t h i s compound. It was known to be of the D-series, and Gorin's study of the r e l a t i v e rates of lead tetraacetate consumption indicated a trans arrangement of the o<-glycol group of Y. Consequently, i t was considered that polyol Y was, i n f a c t , the a l t e r n a t i v e trans_-isomer, 1,5-anhydro-4-deoxy-D-xylo- h e x i t o l ( 6 2 ) , and therefore the enantiomer of Fraction I I , the L-xylo-isomer ( 4 5 ) . .. """" Through the kind cooperation of Dr. Gorin i n supplying a sample of the tris-p-nltrobenzoyl derivative of his polyol, i t was possible to prove t h i s point. Debenzoylation of the derivative with reflu x i n g methanollc sodium methoxide gave the syrupy pol y o l , which was found to havj a s p e c i f i c rotation of + 4 0 ° , equal and opposite to that of the L-isomer ( 4 5 ) , and which had a n.m.r. spectrum i n deuterium oxide solution i d e n t i c a l with that of (45) . The syrupy compound formed a* c r y s t a l l i n e t r i - O - a c e t y l d e r i v a t i v e , m.p. 80-82°, &dj) +40°, whose inf r a r e d spectrum was i d e n t i c a l with that of 2 , 3 , 6 - t r i - - 78 - 0 -a c e ty 1 -1,5 -a nhydr o -4 -deoxy -L-xylo -hex It or, m.p. 80-81°, [ p c ] D _4i°. Comparison of a sample of the c r y s t a l l i n e , levorccta- tory anhydrodeoxyhexitol (Polyol X), also isolated® 2 from the hydrogenolysis reaction of (51), with 1,5-anhydro-4- deoxy-D-lyxo-hexitol (60), one of the products obtained — 122 from the oxo reaction of 3,4-di-O-acetyl-D-arabinal (32) , confirmed Gorin's assignment of the structure of t h i s com- pound. It i s noteworthy that the true s p e c i f i c rotation of each polyol (X and Y) was some 20-30° lower than the values calculated from an application of the rules of o p t i c a l superposition (page 62) . „ (vl) Proton Magnetic Resonance and Stereochemistry of Fractions I and II It was previously noted (page 44) that the resonance positions and i n t e n s i t i e s of the high f i e l d signals a t t r i - buted to the C-4 hydrogens of (44) and (45), provided e v i - dence f o r t h e i r straight chain anhydrodeoxyhexitol structure, rather than the branched chain structures (46) and (47). In the case of one of the isomers, 1,5-anhydro-4-deoxy-L-xylo- h e x i t o l (45), add i t i o n a l stereochemical information could be gained from an inspection of the m u l t i p l i c i t i e s of the signals i n t h i s region, as well as from the resonance po s i - tions of the i n d i v i d u a l C-4 protons, which was, i n complete - 79. ~ agreement with the structure assigned on the basis of the chemical evidence described i n the preceeding pages. Present knowledge of the r e l a t i o n between configura- tions and conformations of carbohydrates and t h e i r nuclear magnetic resonance stems from the pioneering work of Lemieux, 123 K u l l n i g , Bernstein and Schneider . Two factors of impor- tance are the angular dependence of spin-spin coupling con- stants,, and the dependence of chemical s h i f t s on molecular geometry. During t h e i r investigations of various acetylated pyranose sugars, Lemieux and co-workers observed that the s p l i t t i n g of the anomeric protton H^, which i s r e a d i l y d i s - cernable being at lowest f i e l d , was dependent on the re- l a t i v e orientation of Hg, with which i t was coupled. When and Hg were t r a n s - d i a x i a l , the spin-spin coupling constant J w ,=,TT was two--to three'vti'mes larger than when the neighbouring n l ' a 2 hydrogens were i n other •orientations (axial-equatorial or equatorial-equatorial). Thus a large coupling constant was associated with a dihedral angle ( 0 ) of 180° , and a smaller coupling constant with a dihedral angle of 60° . This ex- perimentally observed angular dependence of coupling constants 124 36 was l a t e r generalised by Karplus ' . In the course of the same investi g a t i o n on acetylated 123 pyranose sugars, Lemieux and co-workers observed a d i f - ference i n resonance positions f o r chemically i d e n t i c a l hydro- gens which was related to t h e i r orientations i n space. Thus i n l , 2 , 3 , 4 - t e t r a - 0 - a c e t y l - (5 -D-xylose ( 8 0 ) , the equatorial hydrogen at C-5 was at lower f i e l d than the C-5 a x i a l hydrogen; s i m i l a r l y , equatorial anomeric hydrogens invariably resonated at lower magnetic f i e l d than t h e i r a x i a l counter- parts. The portion of the spectrum of (80) due to the methy- H(X) \ MM) AcO v v " (80) lene hydrogens at the C-5 p o s i t i o n was of p a r t i c u l a r In- te r e s t , as i t was possible to derive parameters for the chemical s h i f t s and spin coupling constants for the i n d i v i d u a l equatorial (A) and a x i a l (B) hydrogens attached to the same 125 carbon atom, when treated as an ABX system . Values ob- tained were: J A B = 12 c/s; J B X = 8 c/s; J A X = 3 .2 c/s. 126 More recently, Woo, Dion and Johnson have made use of the relationships established by Lemieux and co-workers i n deducing the complete configurations of methyl chalcoside ( 8 l ) and of chalcose ( 8 2 ) , degradation products of the a n t i - b i o t i c chalcomycin. Their assignment of the configurations 123 - 81 - at C-3 and C-5 of ( 8 l ) , from the m u l t i p l i c i t i e s of the G-4 proton signals, w i l l be described, as the stereochemistry of the C-3 - C-4 - C-5 fragment of t h i s molecule c l o s e l y resembles the corresponding portion of l ,5-anhydro-4-deoxy- L-xylo-hexitol (45), and therefore provides a model on which to base a discussion of the high f i e l d portion of the spectrum of the l a t t e r compound. With reference to Figure 1A, which shows the signals of the C-4 methylene group of methyl cha l - coside measured i n pyridine s o l u t i o n ; the lower f i e l d group around 8 = 2.05 ppm was assigned to the equatorial hydrogen (H^ e), and the broad group of signals at 8 - 1 .00 to 1.58 ppm, p a r t i a l l y obscured by the C-methyl doublet, to the a x i a l hydrogen (H^ g). Each group of signals was considered 125 • ' •• as an ABX system r e s u l t i n g from coupling between H, , H *+e HQ and one of the neighbouring protons on C-3 or C-5, the re- su l t i n g l i n e s then being further s p l i t by a fourth hydrogen (on C-5 or C - 3 ) . On the basis of the observations of Lemieux and co-workers the spin-spin coupling between the two C-4 - 82 - hydrogens would be expected to be large, of the order of 12 c/s J a dditional coupling of each proton with neighbouring protons on C-3 and C-5 would also be large i f a d i a x i a l re- l a t i o n s h i p existed, otherwise i t would be small. The ob- served s p l i t t i n g of H^e into two quartets indeed showed the anticipated large coupling (12.5 c/s) with the geminal H^a, but gave no additional information on the r e l a t i v e orienta- t i o n of H» and H . However these could be deduced from the 3 5 width of the higher f i e l d H^g s i g n a l , 34.5 c/s, t h i s being p r a c t i c a l l y equal to the sum of three coupling constants ( J 4 a 1^4 J 4 a 3 a n d J 4 a 4e^ w i t h w h i c h t h e a x i a l C-4 hydrogen was coupled to the equatorial C-4 hydrogen and to the two- neighbouring hydrogens on C-3 and C - 5 . Thus, as J ^ a ^ g was 12.5 c/s, J ^ a ^ + J 4 a 5 must be 22 c/s; t h i s large value could only be r a t i o n a l i s e d i f H^ and H^ were both a x i a l . Values of 11 e/s for both J, j - and J., „ s a t i s f a c t o r i l y 4a , d° , _j3 accounted for the observed s p l i t t i n g pattern of H^; the two equ a t o r i a l - a x i a l interactions of H, were assigned J values 4e of 2.1 and $.0 c/s to account for the observed m u l t i p l i c i t y of the lower f i e l d group. The C-4 portion of the n.m.r/ spectrum of 1,5-anhydro- 4-deoxy-L-xylo-hexltol ( 4 5 ) , measured i n deuterium oxide solution (Figure 1 B) i s seen to bear a close resemblance to the corresponding portion of the spectrum of methyl chal- coslde ( 8 l )J the main point of difference i s that the chemical Nmr Spectra (C-4 protons) of : A ' Methyl chalcos ide : B fraction II : C Fract ion I Figure I - 8 3 - s h i f t between a x i a l and equatorial hydrogens at C-4 of (45) i s less than was the case with H^a and H^^ of ( 8 l ), con- sequently there i s no separation between the lower f i e l d and higher f i e l d groups of signals. The s p l i t t i n g of the equatorial hydrogen signal around 8 = 2 . 0 ppm c l e a r l y shows the large spin-spin coupling to the geminal hydrogen (J2j a ,4e approximately 12.5 c/s), and further small s p l i t t i n g s by coupling to the adjacent hydrogens on C-3 and C-5 to give a t o t a l of 8 l i n e s . The width of the signal of the a x i a l hydrogen on C-4 (about 35 c/s) leaves no doubt that and H c are both a x i a l , as the sum of t h e i r coupling constants with H^a Is approximately 22 c/s, as was the case with methyl chalcoside. The corresponding portion of the spectrum of 1 ,5- anhydro-4-deoxy-D~arabino-hexito1 (44) (Figure 1 C) was not amenable to sim i l a r analysis; a multiplet was observed be- tween 8 =1.38 and 2.15 ppm which could not be separated into a x i a l and equatorial signals. Deuterated Analogues of (44) and (45) Several methods are known f o r simplifying, or otherwise modifying, proton magnetic resonance spectra i n order to f a c i l i t a t e t h e i r assignments, and a number of these experi- 127 mental techniques have been discussed by H a l l . As an example of a simple aid to spectral refinement may be - 84 - mentioned the measurement of the spectra of polyols i n deuterium oxide, whereby hydroxyllc hydrogens are exchanged by deuterium. A more v e r s a t i l e , though less r e a d i l y a v a i l - able technique i s the replacement of carbohydrate ri n g hydro- gen atoms by deuterium. Although deuterium has a nuclear spin i t s coupling with adjacent ring hydrogens i s so small that t h e i r signals are merely broadened and show no resolv- able coupling with the deuterium. Consequently, although proton resonance spectra are on the one hand s i m p l i f i e d by the substitution of deuterium f o r hydrogen i n the mole- cule, there i s at the same time a loss of resolution i n the- signals of remaining hydrogens which are adjacent to the. deuterium atoms. This disadvantage can be overcome by the 127 technique of double resonance, or spin decoupling . Very few examples of the use of deuterated analogues as an aid to the assignment of carbohydrate spectra have as yet been 128 129 130 reported ' y i J , and double resonance experiments have been confined to the removal of spin coupling between i n t e r - acting p r o t o n s ^ * 1 ^ 1 . The experiments discussed below, where- by s p e c i f i c deuteration was combined with hydrogen-deuterium decoupling, would therefore appear to be the f i r s t example of t h i s p o t e n t i a l l y powerful technique f o r the s i m p l i f i c a t i o n of n.m.r. spectra. By the substitution of deuterium for hydrogen i n the previously described oxo reaction of 3 , 4-di-O-acetyl-D-xylal - 85 - ( 2 2 ) , It was possible to prepare the deuterated analogues ( (83 ) and (84 ) ) of the D-arabino- (44) and L-xylo- (45) isomers of l ,5-anhydro -4-deoxy-hexitol, i n which one of the hydrogens at C-4 and both of the hydrogens attached at C-6 were replaced by deuterium. In order to reduce the amount of deuterium gas required the experiment was per- formed on a reduced scale, using about 2g of the g l y c a l , and the i n t e r n a l volume of the high pressure bomb was re- duced to approximately 20 ml by the use of a small glass l i n e r , contained i n a hollow metal insert which f i t t e d c l o s e l y inside the bomb. Otherwise, reaction conditions were si m i l a r to those described previously. After removal O H O H (83) (84) .. of c a t a l y s t , and deacetylatlon with methanolic sodium meth- oxide, a syrupy product was obtained whose inf r a r e d spectrum showed absorption i n the region of 2200-2300 cm"1 (C-D st r e t c h i n g ) . Chromatography on paper revealed two main com- ponents with R F values corresponding to those of the normal anhydrodeoxyhexitols (44) and (45)J these were separated on a - 8 6 - preparative scale as described previously. The two chromato- graphlcally pure'fcomponents which were obtained had s p e c i f i c rotations of -11° (R p = 0.46) and -45° (R p = 0.40). The slower-moving, more levorotatory of the two formed a cry- s t a l l i n e acetate with melting point (80-82°) and s p e c i f i c r otation (-42°) very si m i l a r to those of the corresponding derivative of (45). The two fract i o n s were therefore 1,5- a nhydro-4-deoxy-D-a ra b ino-hex i t o1-4,6,6-H2 (83) and 1,5- — 2 anhydro-4-deoxy-L-xylo-hexltol-4,6,6-Hg (84). The n;m.r. spectra of the deuterated isomers, measured i n deuterium oxide solution at 60 Mc/s, showed i n both cases anticipated separation of the ri n g hydrogen signals into a low f i e l d multiplet of r e l a t i v e i n t e n s i t y 6, and a signal at higher f i e l d corresponding to the one hydrogen attached" at C-4, thereby providing add i t i o n a l confirmation that the two components separated from the reaction were (83) and (84). The chemical s h i f t s of the C-4 hydrogens (Figure 2) are of p a r t i c u l a r i n t e r e s t , i n that they can be interpreted as providing information on the mode of addition of carbon monoxide and hydrogen to the double bond of 3,4-di-O-acetyl- D-xylal. It has already been shown, on the basis of the ~~ 123 evidence obtained by Lemieux and co-workers , and also by analogy with the f u l l y assigned spectrum of methyl chalcoside that the equatorial hydrogen at C-4 of 1,5-anhydro-4-deoxy-L- xylo - h e x i t o l (45) resonates around 8 =2.0 ppm, whereas the Oecoupled _L 2.0 1.5 ppm(6) 1 b 5 10 c/s Decoupled CĈ OH 2.0 1.5 ppm (8) Nmr Spectra (H 4) of Deuterated Hexitols Figure 2 - 87 - chemical s h i f t of the a x i a l hydrogen i s about 1.5 ppm. There- fore i t can be assumed that the single C-4 proton signal of ( 8 4 ) , the deuterated analogue of ( 4 5 ) , at about 8 =2.0 ppm (Figure 2A) i s due to an equatorial hydrogen, and consequently the deuterium atom at C-4 i s i n an a x i a l orientation ( 8 5 ) . This assumption i s supported by the fact that the width of the signal at 8 = 2.0 ppm i s only about 9 e/s; a single a x i a l hydrogen at C-4 would be coupled with the two a x i a l hydrogens at C-3 and C -5 , and would therefore have a band width of the order of 20 c/s.. " Most convincing evidence f o r the equatorial orientation of was provided by the deuterium-decoupled spectrum of ( 8 5 ) , also measured i n deuterium oxide s o l u t i o n . Whereas in the normal spectrum (Figure 2A), l i n e s r e s u l t i n g from the coupling of with and Hp.- were broadened by additional coupling with the gem-deuterium atom, and an unresolved 'envelope' was observed, deuterium-hydrogen spin decoupling e f f e c t i v e l y resolved the signal into a sharply-defined quartet - 88 - (Figure 2B), r e s u l t i n g from the s p l i t t i n g of (by H^a or H^g) into a doublet, which was further s p l i t by the other a x i a l hydrogen. The two coupling constants of 2 .3 and 5.1 c/s which f i t t h i s observed s p l i t t i n g pattern could only be accounted for by the fact that was In a gauche re l a t i o n s h i p to the two a x i a l hydrogens at C-3 and C - 5 . Hence the deuterium attached at C-4 and the CDgOH group attached at C-5 were c i s , and the deuterated anhydrodeoxy- h e x l t o l (85) must have been formed by a c i s addition to the double bond of 3 * 4-di - 0-acetyl~I)-xylal. This evidence f o r c i s addition i n the oxo reaction of glycals therefore, supports previous experimental evidence obtained with other, rrentl 17,18 29 30 1 unsaturated compounds , and i s compatible with cu rently acceptable theories of the mechanism of t h i s reaction which were discussed i n the General Introduction.', On t h i s evidence i t was supposed that the isomeric deuterated anhydrodeoxyhexitol (83) must also have the deuterium atom at C-4 and the -CDgOH group at C-5 i n c i s r e l a t i o n s h i p (86) I indeed, the chemical C D O H s h i f t of the single proton at C - 4 , 8 - 1.55 ppm* could well be assigned to an a x i a l hydrogen as i n (87)• However the high f i e l d portion of the spectrum of the normal D-arabino- isomer- ( 4 4 ) , (Figure 1C) , did not permit the assignment of chemical s h i f t s to the i n d i v i d u a l hydrogens at C -4; furthermore, - 89 - the width of the C-4 signal (Figure 2C) i s much less than would be anticipated f o r an a x i a l hydrogen coupled with an a x i a l hydrogen at C-3 and an equatorial hydrogen at C - 5 , as i n (87)• The s p l i t t i n g pattern of the C-4 hydrogen i n (83) which was revealed on hydrogen-deuterium decoupling (Figure 2D) showed a barely resolved t r i p l e t . Thus the coupling constants of with,the two adjacent hydrogens at C-3 and C-5 must be very small, and on the basis of 127 Karplus' parameters t h i s indicated that the dihedral angles between and H , and and were both not f a r removed from 90°• These data were not consistent with a normal chair form (87) f o r ( 8 3 ) . B. Anhydrodeoxyheptitols from 3 , 4 , 6-Tri-CUacetyl-D-galactal In 1957 Rosenthal and Read described the reaction of 3 , 4 , 6-tri-O-acetyl-D-galactal (40) with carbon monoxide and hydrogen under oxo conditions. Following deacetylation of the c a t a l y s t - f r e e reaction product, one compound, m.p. 158- o p —1 _ o 159 » LPdrj+38 , was i s o l a t e d by c e l l u l o s e column chromatography - 90 - which, from i t s empirical formula and those of i t s c r y s t a l l i n e benzoyl and p-nitrobenzoyl derivatives, gave i n d i c a t i o n that a hydroxymethyl group had been added to the double bond of the g l y c a l during the course of the oxo reaction. Prom a study of the oxidation and over-oxidation of t h i s product . 132 by periodate , r e s u l t i n g i n the t o t a l consumption of 6 moles of oxidant and the l i b e r a t i o n of 2 moles of formic acid and one of formaldehyde, i t was concluded that the hydroxymethyl group had probably added at C-2 of the g l y c a l , giving r i s e to the branched chain carbohydrate (88) of un- section comprises a further investigation of the oxo reaction of 3 , 4 , 6-tri-O-acetyl-D-galactal. (i) Reaction Conditions (a) 3 , 4 , 6-Tri-O-acetyl-D-Galaetal (4o) known configuration at C - 2 . The work described i n t h i s AcO (40) (88) The f i r s t recorded preparation of (40) by Levene and 133 "31 Tlpson used the procedure of Fischer- 3 } other modifications - 91 - 134 have been described by Bates and co-workers , and by Overend 42 and co-workers . In t h i s work the convenient procedure 45 of H e l f e r l c h and co-workers was used for the preparation of 3 j 4 , 6-tri-0-acetyl-D-galactai, and one one occasion a good o v e r a l l y i e l d (76$ from D-galactose (18)) of pure g l y c a l was obtained on d i s t i l l a t i o n of the crude reaction product. However, other attempts at p u r i f i c a t i o n , following HQ CH2OH — O QH > H > 0 H Ac 20/H + AcO 2. P/Br 2/H 20 CH 2 OAc — O , > VOAc Zn/AcOH (40) 55 B r OAc (18) (89) preparation by the same procedure, resulted i n rapid de- composition with evolution of acetic acid when the crude product, known to contain a high proportion of the g l y c a l , was heated under vacuum. L i t t l e improvement was obtained 45 when the procedure of H e l f e r i c h was modified by preliminary i s o l a t i o n and p u r i f i c a t i o n of the intermediate 2 , 3 , 4 , 6 - t e t r a - 0-acetyl-oc-D-galactosyl bromide ( 8 9 ) . In cases where d i s t i l l a - t i o n resulted i n decomposition, i t was found, possible to effect a p u r i f i c a t i o n of small amounts of the crude product by chromato: graphy on a column of F l o r i s i l . - 92 - (b) Reaction Conditions and Product I s o l a t i o n Conditions employed for the reaction of 3 , 4 , 6 - t r i - O-acetyl-D-galactal (4o) with 3 moles of synthesis gas (CO + 2H ) were i n general s i m i l a r to those used by 63 Rosenthal and Read , and to those previously described i n Section A for the corresponding reaction of 3 ,4-di-O- acetyl-D-xylal (22). Previous work on the oxo reaction of glycals has indicated that the acetylated hexals are some- what less reactive than the pental derivatives, therefore a s l i g h t l y higher reaction temperature (135°) was used. Reaction products were separated from catalyst and i s o l a t e d as previously described i n Section A, to give a syrupy product whose in f r a r e d spectrum showed that addition to the double bond was complete (absence of G=C stretching vibrations around 1640 cm - 1), and which showed a hydroxyl band of moderate i n t e n s i t y at 3^00 cm - 1. The spectrum was therefore compatible with the presence of one or more sugar alcohols. No attempt was made to fractionate the product i n the acetylated form, as t h i n layer chromatography showed a mixture of components which was not well resolved. The product was then deacetylated with methanollc sodium methoxide 7 2 to give a p a r t i a l l y c r y s t a l l i n e material which was examined by paper chromatography, using the previously described solvent system of water-saturated 1-butanol containing - 93 - 5$ ethanol. When development of the chromatogram was i n - terrupted at the point where the solvent front was near the lower edge, only one major zone of low mobility was apparent on spraying with periodate-Schiff reagent'''3. However on developing the chromatograms for increasingly longer periods the apparently homogeneous main zone became resolved into two d i s t i n c t components, which a f t e r development f o r about 60 hours at room temperature were near the leading edge of the paper and s u f f i c i e n t l y separated to j u s t i f y an attempted f r a c t i o n a t i o n on a preparative scale. Fractionation of a portion of the deacetylated product was car r i e d out by e s s e n t i a l l y the same procedure used to separate the anhydro- deoxyhexitols derived from 3 , 4-di-O-acetyl-D-xylal (Section A). Recovery of the two f r a c t i o n s , which were present i n approximately equal amounts, represented about 75$ of the material applied to the chromatograms. For the purposes of subsequent discussion the faster-moving of the two w i l l be referred to as Fraction A, and the slower as Fraction B. ( i i ) Characterisation of Fractions A and B — 1 1 1 u Rp values of the two fracti o n s were measured on paper with the same solvent system used f o r t h e i r separation, con- firming that each was chromatographically pure. Both f r a c - tions were obtained i n c r y s t a l l i n e form, and gave elemental analyses corresponding to the anticipated empirical formula ) required by the addition of a hydroxymethyl group to the g l y c a l double bond. (a) r -, (b) in. p. R p [ocJ D Fraction A 158-9° 0.24 24° Fraction B 168° 0.21 68° (a) i n water-saturated 1-butanol -f 5$ ethanol at room temperature (b) i n water at room temperature Fractions A and B were both characterised as c r y s t a l - l i n e p-nitrobenzoyl derivatives,- prepared by reaction with 82 p-nitrobenzoyl chloride i n pyridine , whose analyses were compatible with the e s t e r i f i c a t i o n of four hydroxyl groups i n each case. ( i i i ) Structures of Fractions A and B It was possible to determine the molecular structures of Fractions A and B by a s i m i l a r approach to that used i n determining the structures of anhydrodeoxyhexitols I and II (Section A). However, the problem of determining the absolut stereochemistry of the two Fractions was not amenable to solu t i o n by c l a s s i c a l methods. In view of the e a r l i e r Investigation c a r r i e d out by Rosenthal and Read^ 3 on the i d e n t i t y of the single p o l y o l , - 95 - i s o l a t e d from the oxo reaction products of 3 j 4 , 6 - t r l - 0 - a c e t y l - D-galactal, which was thought to have a branched-chain struc- ture (88), i t was of p a r t i c u l a r i n t e r e s t to e s t a b l i s h the configurations of Fractions A and B obtained from the same reaction. As with the hexitols (Fractions I and II) derived from the oxo reaction of 3 , 4-di - 0-acetyl-D-Xylal>-^it was again assumed that four isomeric compounds, (90) -., ( 9 3 ) , were t h e o r e t i c a l possible by the addition of a hydrogen atom and a hydroxymethyl.group at either end of the double bond of (40). Consideration of these four- possible structures CI-tpH HO \—0 CH 2 OH HQ - O X H ^ H CH^DH HQ :.. (90) C H 2 O H C H 2 H - * C — O — C - H I T C H ^ H CH^DH CH 2 OH (91) I C H 2 Q H C H 2 H - C — O - C - H £ H 2 O H i h ^ O H CHJOH (92) 2 C H 2 O H — O , (93) C H ^ H ^ CH 2 OH H - G - O - C H - C - H C ^ O H Ch^OH (94) (95) (96) shows that the argument previously applied i n deciding between branched and straight-chain structures f o r Fractions I and II was equally v a l i d i n - t h i s case* Periodate cleavage * . 9 6 ~ of the oc-glycol group of any one of the four isomeric polyols ( 9 0 ) - ( 9 3 ) and reduction of the dialdehyde so formed would afford a t e t r o l ether which would b e ^ ' o ^ i c f i l l y a ctive, having one asymmetric carbon atom (*) i f formed from one of the straight chain h e p t i t o l s ( 9 0 ) or ( 9 1 ) , but would be devoid of asymmetry and therefore o p t i c a l l y i n - active ( 9 6 ) i f derived from either of the branched Isomers ( 9 2 ) and ( 9 3 ) . When Fractions A and B were separately oxidised with an excess of periodic acid, and the r e s u l t i n g dialdehydes, a f t e r n e u t r a l i s a t i o n of the reaction medium, were reduced i n aqueous solution with sodium borohydride, syrupy products were obtained a f t e r working up the reaction i n the usual way which had equal and opposite rotations; the t e t r o l ether from Fraction A was dextrorotatory ( [pilp + 2 1 ° ) , and that from Fraction B was levorotatory ( [ocj^ - 2 3 ° ) . Confirmation of the structures of the t e t r o l ethers thus obtained was provided by t h e i r n.m.r. spectra, measured In deuterium oxide solution. These were i d e n t i c a l and showed a multiplet at lower f i e l d ( 8 = 3 . 5 0 - 3 . 8 5 ppm) and a group at higher f i e l d {8= 1 . 4 7 - 1 . 0 2 ppm) which had the appearance of a quartet of l i n e s . The r e l a t i v e i n t e n s i t i e s of the two groups of non-hydroxylic hydrogen signals were i n the r a t i o of 1 0 J 2 , consistent with the 2 - d e o x y - 3 - 0 - ( l , 3 - d i h y d r o x y - 2 - p r o p y l ) - g l y c e r o - t e t r i t o l structure - 97 - of t e t r o l ethers (94) and ( 9 5 ) , but not with the structure of the t e t r o l ether ( 9 6 ) , which would have been formed had either Fractions A or B been branched chain structures, as (96) contains only one hydrogen attached to a carbon atom without an oc-oxygen function, and eleven other non- 79 hydroxylic hydrogens which would resonate at lower f i e l d Apart from the r e l a t i v e i n t e n s i t i e s of the low and high f i e l d groups of signals, the close s i m i l a r i t y between the spectra of the t e t r o l ethers obtained from these reactions and those of the enantiomeric 2-deoxy -3-0 - (2-hydroxyethyl)- g l y c e r o - t e t r i t o l s (48) and (49) previously derived from the straight chain anhydrodeoxyhexitols (44) and (45) i s worthy of note. The t e t r o l ethers from Fractions A and B were characterised as the tetra-O-p-nitrobenzoyl derivatives, prepared and Isolated In the usual way. The derivative of o Q Fraction A, m.p. 150-151 , [ 0 ^ + 23 , had an in f r a r e d spec- trum i d e n t i c a l with that prepared from Fraction B, which had m.p. 150-151°, G?^ - 2 3 ° , these data providing further evidence f o r the enantiomeric nature of the two t e t r o l ethers. Apart from the evidence thus obtained from t h e i r de- gradations to o p t i c a l l y active compounds, the n.m.r. spectra of the two polyols A and B (Figure 3) were" also i n d i c a t i v e of the fact that both had straight chain, rather than branched chain structures. Measured i n deuterium oxide solution, the spectra of both A and B showed separation of signals into a FractionI Fraction JT 2X) 1.0 Ppm(S) 4.0 3.0 2.0 J IX) ppm (8) Nmr Spectra of Anhydrodeoxyheptitols F i g u r e 3 - 98 - higher f i e l d group within the region of S = 1.3 - 2.1 ppm (area= 2) and a complex multiplet with a r e l a t i v e area of 8 around 8= 3 .2 - 4 . 3 ppm. The observed i n t e n s i t i e s were compatible only with structures (90) and ( 9 1 ) , as i n D^O solution either of the branched chain polyols (92) and (93) would have only a single hydrogen (cJjH-C) resonating i n the higher f i e l d region, and nine other less shielded hydro- gens. Neither spectrum (Figure 3) showed any spin-spin m u l t i p l i c i t i e s of value i n distinguishing between the two structures (90) and ( 9 1 ) . On t h i s evidence i t could be concluded that both Fraction A and Fraction B were 2 ,6-anhydro -3-deoxy-heptitols, and therefore d i f f e r e d only i n configuration at C - 2 . Thus one of the two fractions was 2,6-a nhydro-3-dteoxy-D-ga la c to - h e p t i t o l (91) , forming the t e t r o l ether 2-deoxy - 3 - 0 - ( 1 , 3 - dihydroxy - 2-propyl)-L-glycero-tetritol (95) on cleavage with periodate and reduction with sodium borohydride, and the other f r a c t i o n was the corresponding D-talo-lsomer ( 9 0 ) , from which the t e t r o l ether (94) having the D-configuration was obtained. - 9 9 " H' i _ 0 - C H , (94) (91) (90) (iv) Stereochemistry of Fractions A and B (a) Periodate Consumption and Acetalation Determination of the amount of periodate ion con- i sumed by both Fractions A and B, by the previously described 76 • • | spectrophotometry method , confirmed the presence i n each periodate ion reacting i n each case. The rates of reaction with periodate ion r e l a t i v e to those previously determined for the two anhydrodeoxyhexitols (44) and (45) were of i n - te r e s t , the reaction being much fas t e r f o r the two anhydro- deoxyheptitols A and B than f o r (44) and ( 4 5 ) , i n which a trans arrangement of secondary hydroxyl groups was present. This i s c l e a r l y shown i n the table on the next page, which compares the mole f r a c t i o n of periodate ion consumed against time f o r anhydrodeoxyhexitol (45) and anhydrodeoxyheptitol B, compound of one cx-glycol group, one molar equivalent of - 100 - which were reacted under Identical conditions. Time 1 min. 47 min. 1 .5 hrs. 3 hrs. 10.5 hrs. 17 hrs. 24 hrs. (a) 00 O.63 0.97 0.40 0.59 0.95 0.87 0.88 Moles of Periodate Ion Consumed per Mole of Substrate f o r (a) 2 ,6-Anhydro -3-deoxy-heptitol (Fraction B)J (b) 1,5-Anhydro«4-deoxy-L-xylo-hexitol ( 4 5 ) . The exact mechanism by which the carbon-carbon bond connecting the adjacent hydroxyl groups of an oc-glycol i s broken has not been established conclusively, but a c y c l i c ester structure such as (97), f i r s t proposed by Criegee 1"^, i s generally considered to be ah intermediate i n the proc- 74 ess . The well known difference i n rate of periodate -C—0. -C—Or : I O 4 H 3 (97) oxidation between c i s and trans oc-glycol groups attached I36 to a six-membered,ring has been r a t i o n a l i s e d oh con- 137 formatiohal grounds by Hoheymah and Shaw . Variations observed i n the rate of periodate consumption of a number of pyranoside derivatives, i n which only one oc-glycol group - 101 - of known conformation was available f o r oxidation, were con- sistent with the fact that the rate was dependent on the r e l a t i v e ease of formation of a c y c l i c intermediate such as ( 9 7 ) . Attachment of such a 5-membered c y c l i c structure onto a pyranoid r i n g would require the C-0 bonds of the oc-glycol group to be constrained into a greater degree of . coplanarity for both c i s - ( a x i a l - e q u a t o r i a l ) ( 9 8 ) and trans- (equatorial-equatorial): ( 9 9 ) d i o l s . In the l a t t e r case ( 9 9 ) , as the two carbon atoms of the d i o l systiti are rotated i n order to bring the two attached oxygen atoms closer to- gether, the r i n g becomes more puckered; consequently the...axial a x i a l substituents on the r i n g are brought closer together and considerable energy i s required to overcome the r e s u l t i n g repulsions. The reverse applies with c i s (axial-equatorial) systems ( 9 8 ) , where a s i m i l a r operation r e s u l t s i n the r i n g becoming less puckered, and interatomic repulsions are not appreciably changed. Formation of a c y c l i c intermediate there- fore proceeds r e a d i l y . The rapid rates of periodate oxidation (98) (99) - 102 - of Fractions A and B r e l a t i v e to those of anhydrodeoxyhexitols (44) and (45) thus demonstrated the c i s configuration of the secondary hydroxyl groups at C-4 and C-5 of the two anhydro- deoxyheptitols. A s i m i l a r explanation f o r the more ready formation of the O-isopropylidene derivatives of an oc-glycol group situated on a six-membered ring when the hydroxyl groups are c i s , rather than trans, has been proposed by Angyal and 138 Macdoiald . Here the difference i n energy required to assume a greater degree of coplanarity for the —<jj- groups of (98) and (99) i s such that under normal conditions the O-isopropylidene derivative of a trans d i o l does not form. It should be noted, however, that complete coplanarity of the f i v e atoms of the 1 ,3-dioxolane type r i n g of O-isopropylidene derivatives i s not, as was once supposed, necessary for t h e i r TOQ nil© formation . Additional evidence f o r the presence of two adjacent secondary hydroxyl groups having a c i s arrange- ment i n one of the anhydrodeoxyheptltols was provided by i t s ready reaction with acetone, subsequent evidence showing that the mono-O-isopropylidene derivative so formed did not gngage either of the primary hydroxyl groups i n the molecule. The r e s u l t s obtained could be interpreted i n terras of the possible stereochemistry at C-2 of 2,6-#§taydro-3-deoxy-heptitol B. - 103 - The reaction of Fraction B with acetone at room tempera- ture, catalysed by a trace of sulphuric acid, gave an o i l from which the reaction product was r e a d i l y removed by extraction . with b o i l i n g carbon t e t r a c h l o r i d e , i n which the unreacted polyol was insoluble. The l a t t e r could then be subjected to further treatment with a c i d i f i e d acetone, thereby enabling the product to be obtained i n good o v e r a l l y i e l d . The cry- s t a l l i n e derivative, m.p. 104-105°, [pcj-p + 12°, gave an e l e - mental analysis i n agreement with the emp|fp.eal formula (G _H,oO ) required f o r the mono-O-isopropylidene derivative (100) of a 2 , 6-anhydro - 3-deoxy-heptitol. The presence of two free hydroxyl groups was shown by conversion of (100) (100) (102) to a di-0-p4;olysulWi°nyl derivative (101) , m.p. 135°>. • -42°. It was then possible to confirm that both hydroxy!' groups of the mono-O-isopropylidene derivative (100) not involved i n acetal formation were primary. When the di-O-p-tolysulphonyl derivative (101) was heated i n a sealed tube with sodium iodide - .104 - i n acetone solution f o r 26 hours at 1 1 8 ° , the amount of sodium p-toluenesulphonate which was subsequently i s o l a t e d by f i l t r a t i o n was equivalent to the replacement of two tbsy- loxy groups by iodide, to give the 1 ,7-dideoxy- derivative (102) (which was not obtained i n 1 s u f f i c i e n t amount for chara- ct e r i s a t i o n ) . It i s well known that primary tosyloxy groups are replaced by iodide under these conditions, whereas those 94 at secondary positions are generally unreactive . These reactions therefore demonstrated the presence of two adjacent secondary hydroxyl groups having a c i s r e l a t i o n s h i p , i n addi- t i o n to two primary hydroxyls, i n Fraction B. . In order to obtain the separation of an amount of sodium p-toluenesulphonate equivalent to the replacement of both primary tosyloxy groups of (101) , i t was necessary to employ a longer time and a higher temperature than i s usual f o r t h i s reaction. However, the formation of an appreciable amount of sodium p-toluenesulphOHate was !bbL served within minutes of the sta r t of the reaction, while the temperature was s t i l l below 100°, and i n a separate ex- periment c a r r i e d out under the same conditions, but at a temperature of 100°, sodium p - t o l u e n e s u l ^ ^ ^ ^ a f t e r 25 minutes equivalent to the replacement of one tosyloxy group by iodide. It was apparent therefore, that whereas one of the primary tosyloxy groups of (101) was replaced by iodide with unusual ease, the other was replaced with d i f f i c u l t y . - 105 - I t has been shown that the proximity of an acetal ring can render an otherwise reactive primary tosyloxy group rather i n e r t to the action of sodium iodide. Thus whereas 2 , 6 - d i - 0 - p-tolyls\tlphonyl-D-ga lactose was r e a d i l y converted to the 6-deoxy-6-iodo- derivative i n good y i e l d , the reaction of the corresponding 3>4-0_-isopropylidene derivative (103) with sodium iodide under i d e n t i c a l conditions resulted i n only 18$ replacement of the primary tosyloxy group by iodide ; S i m i l a r l y , replacement of the ^primary tosyloxy group i n 1 ,2! (101) 3,4-di - 0-isopropylidene - 6 - 0-p-tolylsulphonyl- oc-D-galactose (104) requires heating with sodium iodide i n acetone for 118 36 hours at 125° . The s t r u c t u r a l s i m i l a r i t y of (101) to (103) and (104) i s apparent; hence, by analogy, i t may be concluded that, of the two primary tosyloxy groups of (101) , that at 0 - 7 , adjacent to the acetal r i n g , was the one which reacted slowly with sodium iodide at 1 1 8 ° , and o that the C - l tosyloxy group reacted very r e a d i l y at 100 . - 106 - I t was then of Interest to speculate on the l i k e l y orientation of thejreactive terminal tosyloxy group of (101), which must be either equatorial (in 2 , 6-anhydro - 3-deoxy - 4 , 5-Q-lsopropyll- dene-l ,7-di -0-p-tolysulphonyl-D-gaJLaeto_-heptitol (1^05)) or a x i a l (in the corresponding derivative (106) having the taj|6- configuration). Evidence i s available which demonstrates that an axially-disposed primary tosyloxy group w i l l react with sodium iodide with greater ease than an equatorial group, when both -i 42 are i n s i m i l a r environments . The b i c y c l i c diacetals of he x i t o l s , by c i s fusion of ifao six-membered rings, give structures possessing two terminal hydroxymethyl groups which can be a x i a l or equatorial, depending on the configura- t i o n of the parent h e x i t o l . Thus 2,4:3,5-di-0-methylene-D- 128 ~~ g l u c i t o l (107) , i n the preferred conformation , has the terminal C-6 group a x i a l and the C - l group equatorial. I t has been shown that, on reaction of the l , 6 - d i - 0 - p - t o l y l - sulphonyl derivative of (107) with sodium iodide, the a x i a l - 107 - (107) tosyloxy group at C-6 was replaced more r e a d i l y than the 143 equatorial C - l group . S i m i l a r l y , the replacement by iodide of the tosyloxy groups of 2,4:3,5-di-0-methylene- 1,6-dl-O-p-tolylsulphonyl-L-idltol (both equatorial) was ol44 ~~ very slow at 100 , whereas with the corresponding d e r i - vative of D-mannitol, both (axial) tosyloxy groups reacted 145~ r a p i d l y . On the basis of these data, i t was considered that (101) may have had the 0-1 tosyloxy group i n an a x i a l ori e n t a t i o n ( 106) , i n which case Fraction B would be 2 , 6 - anhydro-3-deoxy-B-talo-heptitol ( 9 0 ) . Subsequent evidence showed that t h i s speculation was, i n fact,, correct. . It would have been of in t e r e s t to compare the same reactions of the corresponding derivatives of Fraction Aj however, t h i s compound, being sparingly soluble In acetone, did not form an O-isopropylidene derivative In s u f f i c i e n t amounts for characterisation and d e r i v a t i s a t i o n . . 1 - 108 !- (b) Stereochemistry at 0-2 Attempts were made to solve the problem of the con figurations of the enantiomeric t e t r o l ethers obtained from'Fractions A and B, (and hence the absolute stereo- chemistry of the two anhydrodeoxyheptitols), by the prepara- t i o n of one.of them from a suitable structure whose configura- t i o n at the p o t e n t i a l asymmetric centre (0-3 of the t e t r o l ethers) was known (a procedure which had previously been successful i n proving the structures of the anhydrodeoxy- hexitols (44) and (45) by c o r r e l a t i o n with ( 5 9 ) ) . By the same approach, possible structures which would afford a 2-deoxy - 3 - 0-(l , 3-dihydroxy - 2-propyl)-glycero-tetritol (109) were! (a) 2 ,6-anhydro -3-deoxy-heptitols (108) other than Fractions A and B, and (b) 3 ,6-anhydro -2-deoxy-heptitols (11©), both by periodate cleavage of the C-4 - C-5 bond and sub- sequent reduction. At the time t h i s i n v e s t i g a t i o n was CHPH 59 C H O H 2. (108) (109) ( no ) I - 109 - undertaken no other compounds of the former type (108) were available whose stereochemistry had been established, and a search of the l i t e r a t u r e f a i l e d to reveal any known 3 , 6 - anhydro-2-deoxy-heptitols (Hip). However, i t was considered that a possible synthetic route to the l a t t e r type of structure (110) would be the a p p l i c a t i o n to a 2,5-anhydro- hexose ( i l l ) of a known procedure (the Fischer-Sowden 146, nitromethane synthesis ) for the preparation of homolo- gous 2-deoxy-aldoses, the product (112) on reduction then affording the required 3 ,6-anhydro -2-deoxy-heptitol ( l l O ) , whose configuration at C-3 would be the same as C-2 of ( i l l ) J thus OHO CHO A H 2 C h W . A h l - ^ I \ Fischer- 1, A Reduction. I O H 1 60 CHOH\ • >GHOH\ — > (no) > (io9) CHOHy S o w d e n GHOH/ inpH AH^OH (111) (112) (°) Attempted Syntheses of 3,6-Anhdyro-2-deoxy-heptltols (110) Of the 2,5-anhydro-hexoses, the best known i s 2 , 5 - anhydro-D-mannose (chitose) ( 1 1 5 ) > an amorphous material, 147 2,5-anhydro-D-mannose has been known fo r many years 1, but only comparatively recently has i t s structure been established - 110 - 148,149 beyond doubt ' , confirming the previous conclusions of Levlne and LaForge . The formation of (115) by the nitrous acid deamlnatlon of 2-amino-2-deoxy-D-glucose (113) involves an inversion i n configuration at C-2, and i s con- sidered to proceed via an intermediate diazonium s a l t 95 (114) J subsequent elimination i s accompanied by rearward attack of the ring oxygen at C-2, r e s u l t i n g i n rearrangement of a s i x - to a five-membered r i n g . 2,5-Anhydro-D-glucose 6 l CHO (115) (117), the eplmer of (115).,- i s also known and has been ob- tained by the analogous deamlnatlon of 2-amino-2-deoxy-D- mannose hydrochloride (.116) by the action of mercuric 151 oxide""''"", again with inversion at C-2 CHpH HCI CHf)H H,OH 62 (116) (117) - i i i - Though l e s s - r e a d i l y a v a i l a b l e , 2,5-anhydro-D-glucose (117) was selected as being a more suitable s t a r t i n g material f o r the attempted preparation of a 3,6-anhydro-2-deoxy- h e p t i t o l (110) than i t s epimer (115) for two reasons: (a) i t i s reported to be a h i g h l y - c r y s t a l l i n e s o l i d , whereas (115) i s an i l l - d e f i n e d syrup, and (b) the pre- paration of i t s precursor, 2-amino-2-deoxy-D-mannose hydro- chloride ( l l 6 ) from D-arabinose (118), according to Equation 6 3 , would afford a series of intermediate compounds ( ( 119) , (120) and (121)) which would provide useful models f o r following the subsequently planned conversion of (117) to 3,6-anhydro-2-deoxy-D-manno-heptltol (124), also via the nitre-methane synthesis for 2-deoxy-sugars (Equation 64). 152 The preparation, described by Sowden and Oftedahl of 2-amino-2-deoxy-D-mannose hydrochloride ( l l 6 ) from D- arabinose (118), by way of D-arabino-tetraacetoxy-l-nitro- 1-hexene (121) (Equation 63) proceeded smoothly, with forma- t i o n of h i g h l y - c r y s t a l l i n e intermediates. However, the attempted conversion of ( l l 6 ) to 2,5-anhydro-D-glucose (117), 151 by the method of Levine involving the action of mercuric oxide, did not give the anticipated c r y s t a l l i n e anhydro- sugar (117)5 an amorphous brown s o l i d was obtained which could not be p u r i f i e d . - 112 - Attention was then turned to the use of the more accessible 2,5-anhydro-D-mannose (115); t h i s was made by 15 3 — the method of Grant , who investigated the s t a b i l i t y of C H O HO-C-H H-C-OH" H-C-OH 4 (118) H 2 0H " CH 2N0 2 CHJM H-C-OH HO-C-H HO-C-H- HO-C-H + H - C - O H H - C - O H - O H O H H O H 2 A Ĥ OH (119) h^N0 2 C K N O ^ H-C-OAc A c O - C - H A c O - C - H H - C - O A c f A c O - C - H hkC-OAc H - C - O A c H-C-OAc i h ^ O A c (120) wHOAc CH I AcO-C-H * H-C-OAr" H-i-OAc CH^OAc (121) CH^NC^ H-C-NH-€Ac H O - C - H H - C - O H H - A - O H A (123) HOH 2 A: 'r-HN-C—H H O - C - H + I H - C - O H H - C - O H C 63 HPH (122) (116) I (117) (117) CHpH — Q HO O H HO CHpH — O . O H C H O H CH^OH — O OH HO 64 W H (124) 9 1 CH-NO, 6l H^OH - 1 1 3 146 ( 1 1 5 ) under various conditions, and established optimum conditions f o r i t s preparation i n a reasonably high state of purity. This was confirmed by the formation of a con- siderable y i e l d of the p-nitrophenylhydrazone derivative of ( . 1 1 5 ) from the syrupy product. When the anhydro-sugar ( 1 1 5 ) thus prepared was reacted with nitromethane i n the. presence of sodium methoxide under anhydrous conditions, according to the general procedure of Fischer and SowdenJ , the immediate formation of a copious amount of white s o l i d was observed, presumed to be the anticipated aci-sodium s a l t s of nitroalcohols ( 1 2 5 ) and ( 1 2 6 ) . Following de- i o n i s a t i o n by passage of an aqueous solution of the product through Dow ex 50(H-') r e s i n , a syrup was obtained whose i n - frared spectrum (Figure 4A) c l o s e l y resembled that (Figure 4B) of the mixture ( . 1 1 9 ) of 1-deoxy-l-nitro-D-mannitol and l-d e o x y - l - n i t r o - D - g l u c i t o l previously obtained during the preparation of 2-amino~2~deo:xy-D-mannose hydrochloride' ( 1 1 6 ) (Equation 6 3 ) . Infrared evidence, and the fact that th i n layer chromatography of the product showed i t to con- s i s t e s s e n t i a l l y of two closely-moving components, indicated that the f i r s t stage of the attempted synthesis of 3,6-anhydro- ( 1 1 5 ) QH20H —O (pH + -C-CHNO H 0 I 2 2 H OH •c- l OH - C - C H N Q I 2 2 6 5 ( 1 2 5 ) ( 1 2 6 ) Infrared Spectra of: A and C - nitroalcohols and acetylated nitroalcohols from 2,5-anhydro-D-mannose B and D - nitroalcohols and acetylated nitroalcohols from D-arabinose Figure 4 - ilii - 2-deoxy-D-gluco-heptltol had gone as planned. Acetylatlon of the product with acetic anhydride containing a trace of sulphuric acid gave a mixture of acetates whose inf r a r e d spectrum (Figure 4C) also c l o s e l y resembled that of the mixture (120) previously obtained by the acetylatlon of (119) (Figure 4D), showing absorption c h a r a c t e r i s t i c of N 0 g stretching vibrations at 155© and 1365 cm"1. When the acetylated material was refluxed i n benzene solution with sodium hydrogen carbonate, the inf r a r e d spectrum of the product which was is o l a t e d was v i r t u a l l y unchanged from that of the acetylated reactant. Elimination of acetic acid, to give the required n i t r o - o l e f i n , on ; .. treatment with mild base would presumably have been c l e a r l y apparent by a s h i f t to lower frequencies of the asymmetric and symmetric stretching vibrations of the NOg gjpoup on 154 conjugation , as was evidenced by the two bands at 1510 -1 -1 cm and 1350 cm i n the infrared spectrum of D-arabino- tetraacetoxy-l-nitro-l-hexene (121). Prolonged r e f l u x i n g i n benzene solution with sodium hydrogen carbonate, and sim i l a r treatment with the stronger base sodium acetate, f a i l e d to bring about the required elimination. This approach to the synthesis of a 3 ,6-anhydro -2-deoxy-heptitol, which apparently showed some promise i n the e a r l i e r stages, was therefore abandoned as an alt e r n a t i v e s t r u c t u r a l proof, de- scribed below, became a v a i l a b l e . - 115 - (d) Correlation with 2 ,6-Anhydro-3-deoxy-D-gluco-heptltol (130) It was subsequently possible to e s t a b l i s h the stereo- chemistry of one of the enantiomeric t e t r o l ethers previously described, and hence the absolute configurations of both Fractions A and B,/by c o r r e l a t i o n with the t e t r o l ether derived from a 2 ,6-anhydro -3-deoxy-heptitol (108) of known 155 stereochemistry. Rosenthal and Koch have investigated the oxo reaction of the commercially-available 3 , 4 , 6 - t r i - O-acetyl-D-glucal (20 a). The products from t h i s reaction were found to be completely analogous to those obtained from the reactions of 3 , 4 , 6-tri-O-acetyl-D-galactal (40) and 3 ,4-di-O-acetyl-JD-xylal ( 2 2 ) . When the ca t a l y s t - f r e e mixture of isomeric 4 , 5 , 7-tri - 0-acetyl - 2 , 6-anhydro - 3-deoxy- h e p t i t o l s (127) and (128) r e s u l t i n g from the oxo reaction of (20 a) was reacted with p-bromobenzenesulphonyl bromide i n pyridine, one of the isomers, (127) , was r e a d i l y i s o l a t e d as the c r y s t a l l i n e O-p-bromobenzenesulphonate, X-ray 156 analysis of the c r y s t a l l i n e derivative established that i t was the 1-0-p-bromobenzenesulphonyl derivative (129) °f 4 , 5 , 7 - t r i - 0 - a c ety1-2 ,6-a nhydro -3-deoxy-D-gluco-heptitol (127) . Deacetylation of the ca t a l y s t - f r e e oxo product and fr a c t i o n a t i o n by paper chromatography gave two isomeric anhydrodeoxyheptitols. One of these, 2,6-anhydro-3-deoxy- D-gluco-heptitol (13©) was also obtained from the 1-0-p- bromobenzenesulphonyl derivative (129) , of known configuration, - Il6 - "by standard procedures C H z O A c AcO OAc (20a) ci-yoAc —O^HpBs OAc AcO (129) AcO CH,OAc ^ O A c — O v C h ^ O H . + A , 0 OAc Ch^OH (128) (130) 2,6-Anhdyro-3-deoxy-D-gluco-hept11o1 (130) was subjected to periodate oxidation and the r e s u l t i n g dialde- hyde was reduced with sodium borohydride, using the pro- cedure described previously, to afford 2-deoxy - 3 - 0 - ( 1 , 3 - dihydroxy - 2-propyl)-L-glycero-tetritol ( 9 5 ) . This was dextrorotatory, ( [p^j) + 2 6 ° ) , as was the t e t r o l ether ;H 2 OH H C — O - A HPH (95) Fraction A 6' (130) from Fraction AJ (95) formed a tetra-O-p-nitrobenzoyl d e r i - vative, m.p. 151-152 c , whose melting point was not depressed on admixture with the corresponding derivative of the t e t r o l ether from Fraction A, and the two p-nitrobenzoates had the same s p e c i f i c rotations and i n f r a r e d spectra. Therefore Fraction A must have the B-configuration at C - 2 , as had the B-gluco-isomer ( 130) , t h i s being the only asymmetric centre surviving i n the t e t r o l ether (95) formed by cleav- age of the C-4 - C-5 bond and reduction. As i t could be safely assumed that the secondary hydroxyl groups at C-4 and C-5 of the anhydrodeoxyheptitols were unchanged i n configuration from those of B-galactal, the structure of Fraction A was therefore established as 2,6-anhydro-3-deoxy- B-galacto-hept11o1 (91) . As Fraction B d i f f e r e d from Fraction A only i n configuration at C - 2 , t h i s must be 2 , 6 - anhydro - 3-deoxy-B-talo-heptitol ( 9 0 ) . (v) S t e r i c Aspects of Hydroxymethylation The reactions of acetylated glycals with carbon monoxide and hydrogen to give, as major products, an approximately equal r a t i o of the two possible products re- s u l t i n g from addition of a hydroxymethyl group at G-l of the g l y c a l would appear to be general; t h i s i s supported by the fact that e s s e n t i a l l y s i m i l a r r e s u l t s are obtained by the oxo reaction of 3 , 4 , 6-tri-O-acetyl-B-glucal (20 a ) 1 - ^ and 3,4-di-O -acetyl-B-arabinal ( 3 2 ) 1 2 2 . That hydroxymethylation - 118 - occurs at the carbon of the double bond (X to the r i n g oxygen of the glycals Is consistent with the findings of other workers regarding the appli c a t i o n of the oxo reaction to 25 c y c l i c v i n y l ethers such as 2 , 3-dihydro - 4 H-pyran (l) . . 26 and furan (9) ' . However, i n view of the strong s t e r i c dependence of the oxo reaction when applied to o l e f i n s i n general, and the apparent s t e r i c factors involved i n addi- t i o n reactions of g l y c a l s , both of which were discussed i n the General Introduction, i t i s surprising that approxi- mately equal amounts of two straight-chain products are always obtained, as a r e s u l t of c i s addition to both sides of the g l y c a l double bond, rather than a preponderance of the isomer which would r e s u l t from addition at the l e a s t - hindered side. 17 Heck and Breslow found that, at oxo reaction temperatures, the formation of alkylcobalt tetracarbonyls from o l e f i n and cobalt hydrotetracarbonyl (Equations 7 - 9 ) , i n what i s assumed to be the f i r s t stage of'the oxo reaction, i s r a p i d l y r e v e r s i b l e . These workers found that the product d i s t r i b u t i o n obtained by reaction of o l e f i n s with cobalt hydrotetracarbonyl at low temperature was often not the same as that, found under normal oxo conditions: f o r example, whereas isovaleraldehyde i s the major product from the oxo reaction of isobutylene (equation 15), at low temperature trimethylacetaldehyde, r e s u l t i n g from addition to the more - 119 - highly substituted side of the double bond, predominates. These findings were explained by the fact that, as the i n i t i a l addition to the double bond i s r e v e r s i b l e , the product d i s t r i b u t i o n at high temperature i s a r e f l e c t i o n of the r e l a t i v e s t a b i l i t i e s of the adducts, rather than t h e i r i n i t i a l concentrations, the terminal aldehyde being more stable yet forming less r e a d i l y . A s i m i l a r explanation could account for the observed isomer d i s t r i b u t i o n from the oxo reaction of g l y c a l s . Assuming that addition to the least-hindered side of the double bond i s favoured, and that the f i r s t stage of the reaction leads to glycosylcobalt tetracarbonyls .according to the generally accepted mechanism, i t would be reasonable to suppose that the adduct so formed (e.g. (131) from 3 , 4 - di-O-acetyl-D-xylal) would be less stable than the alterna- t i v e , k i n e t i c a l l y less-favoured adduct (132) with the -Co(CO)^ group i n equatorial orient a tlonj|and would revert more r e a d i l y to the g l y c a l . AcO- •Co(£0)4 (131) (132) - 120 - C. Hydroformylatlon of 3 , 4-Di-O-acetyl-D-xylal In section A i t was shown that the reaction of 3 , 4 - di-O-acetyl-D-xylal with 3 moles of synthesis gas gave a mixture consiting predominantly of the 2 , 3 - d l - 0 - a c e t y l - l , 5 - anhydro-4-deoxy-hexltols (74) and ( 7 5 ) , from which (44) and (45) were obtained on deacetylation. An a_ p r i o r i assumption regarding the oxo reaction of glycals i s that the i n i t i a l products are aldehydes, r e s u l t i n g from the hydroformylatlon of the double bond, and that these sub- sequently undergo hydrogenation to afford the observed products. On t h i s basis i t was reasonable to assume that termination of the reaction at the point where only 2 moles of synthesis gas had reacted would give a product consisting predominantly of aldehydes; f o r example the anticipated products from the hydroformylatlon of 3 , 4-di-O-acetyl-D- x y l a l (22) would be the isomeric 4 , 5 - d i - 0 - a c e t y l - 2 , 6 - anhydro-3-deoxy-aldehydo-hexoses (133) and ( 134) . - 121 - A ce r t a i n amount of evidence has been obtained pre- viously to show that aldehydlrj compounds are present i n products obtained by the oxo reaction of g l y c a l s ; f o r example, t h i s has been indicated by the reducing power of reaction products to Somogyi's reagent 1-^. From the reaction of 3*4,6-tri- x ' 64 0-acety1-D-gluca1 (20 a) with carbon monoxide and hydrogen , a single c r y s t a l l i n e component was Isolated, i n addition to the acetylated alcohols (127) and (128), which was though to be an O-acetylated-anhydrodeoxy-aldehydo-heptose, giving (130) on deacetylation and reduction. Additional evidence f o r the presence of an aldehydo-compound i n the hydroformy- l a t i o n product of (20 a) has been o b t a i n e d 1 ^ 0 n i t s reac- t i o n with ethyl mercaptan, as a homogeneous syrup was i s o l a t e d by chromatography whose n.m.r. spectrum was con- sistent with an acetylated t h i o a c e t a l . ( i ) Reaction. Conditions and.Product I s o l a t i o n Conditions for the reaction of 3,4-di-O-acetyl-D- x y l a l with 2 moles of synthesis gas ((22) >(133) + (134), Equation 68) were s i m i l a r to those previously described for i t s hydroxymethylation (Section A), except that a lower temperature (115°) was employed f o r the reaction, to favour the formation of aldehydes. Once absorption of synthesis gas commences, as evidenced by a decrease i n the pressure indicated on the guage attached to the reaction vessel, i t r a p i d l y overshoots the hydroformylation stagehand i t was - 122 - necessary to carry out a preliminary experiment under Iden- t i c a l conditions i n order to determine the point equivalent to the absorption of 2 moles of gas, when the reaction was quenched by immersion of the reaction vessel In i c e . The reaction product was i s o l a t e d i n a s i m i l a r manner to that described previously. Catalyst was re- moved from the reaction mixture by elu t i o n from a F l o r i s i l column with petroleum ether. The f r a c t i o n which was then eluted with benzene-isopropyl alcohol weighed only 10.0 g (from 12.0 g of (22)) a f t e r evaporation of solvent and therefore did not represent the whole of the reaction pro- duet. A further 2.5 g of a mobile syrup was i s o l a t e d from the petroleum ether eluate; t h i s c r y s t a l l i s e d on standing, and was i d e n t i f i e d as 3,4-di-O-acetyl-D-xylal by compari- son of i t s thin layer chroraatogram and i n f r a r e d spectrum against an authentic specimen of the g l y c a l . It was ap- parent, on t h i n layer chromatography of the main f r a c t i o n eluted from F l o r i s i l with benzene-isopropyl alcohol, that i t also contained an appreciable quantity of the unreacted g l y c a l , t h i s being further indicated by a sharp peak at 1640 cm 1 i n the i n f r a r e d spectrum of t h i s f r a c t i o n . In addition to 3,4-di-O-acetyl-D-xylal, t h i n layer chromato- graphy revealed the presence of a compact mixture of other components, (closely resembling i n mobility and general appearance the mixture r e s u l t i n g from the hydroxymethylation of (22) (Section A)). This mixture of reaction products, being well separated from the faster-moving g l y c a l , was r e a d i l y i s o l a t e d by chromatography onj a column of alumina. A portion of the main f r a c t i o n thus separated gave 3>4- di - 0_-acetyl-D-xylal and reaction products i n the r a t i o of 2 : 3 . Taking into account the add i t i o n a l amount of g l y c a l eluted from P l o r i s i l along with the c a t a l y s t , the composition of the mixture obtained from the reaction was therefore approximately 50$ unreacted 3 * 4-di - 0-acetyl-D- x y l a l and an equal amount of saturated products. Comparison of the reaction product, separated from unreacted g l y c a l as described above, with the product pre- viously obtained by the reaction of (22) with 3 moles of synthesis gas (Section A), showed that the two were p r a c t i - c a l l y indistinguishable on the basis of t h i n layer chroma- tograms and Infrared spectra. However a difference was found i n the low f i e l d region of the n.m.r. spectra, as only the product i s o l a t e d from the above reaction showed a single, peak at & =9.35 ppm, c h a r a c t e r i s t i c of aldehydic IRQ protons . On the assumption that t h i s was due to the presence of compounds ((133) + (134)), and that the remainder of the mixture comprised t h e i r reduction products ((74) + (75))> the i n t e n s i t y of the low f i e l d absorption r e l a t i v e to the combined i n t e n s i t i e s of the acetoxy-hydrogen signals i n the region of S - 2 .0 - 2.2 ppm indicated that approxi- mately 20$ of aldehydo-compounds were present i n the mixture - 12.4 - Results of t h i n layer chromatography c l e a r l y precluded the p o s s i b i l i t y of e f f e c t i n g a separation of the mixture by chromatographic procedures. ( i i ) Reac tlo.n^.-lth;^ 2i4^B-ihitrophenyl.hydra zine Confirmation of the presence of aldehydie components i n the main f r a c t i o n from the attempted hydroformylatlon reaction of 3 , 4-di-O-acetyl-D-xylal was sought by reacting a portion of the product with a test solution of 2 , 4 - d i n i t r o - ii • 160 phenylhydrazine, which was prepar/e|3, i n the usual way , \ by d i s s o l v i n g the reagent i n aqueous sulphuric acid and d i - l u t i n g with ethanol. Immediate p r e c i p i t a t i o n of a yellow s o l i d was indeed observed; t h i s changed to an orange o i l on standing. However, a control experiment i n which pure 3 , 4-di-O-acetyl-D-xylal, known to be present i n the hydro-, formylation product, was separately treated i n the same way also resulted i n the p r e c i p i t a t i o n of an or&nge s o l i d i n considerable amount. C l e a r l y , under these conditions, i s o l a t i o n of any derivatives r e s u l t i n g from the reaction with 2 , 4-dinitrophenylhydrazine of aldehydo-hexoses present i n the mixture would have been corapiicated^^y;;'fehe;--presence of derivatives o r i g i n a t i n g from the g l y c a l . The nature of the l a t t e r product was not investigated further, but i s presumed to r e s u l t from one or both of two known reactions of glycals under aqueous acid conditions, namely the addition of water across the double bond leading to 2-deoxy-aldoses , l 6 l and t h e i r ready isoraerisation to pseudo-glycals . Addi- t i o n a l complications such as deacetylation would be a n t i c i - pated under these strongly acid conditions, indicated by the observed change from a s o l i d to an o i l . Conditions were therefore established f o r reacting the hydroformylation product with 2,4-dinitrophenylhydrazine i n the absence of strong a c i d . A solution of the reagent i n ethanol, saturated at the b o i l i n g point, was added i n portions to a solution of the reaction product i n ethanol containing a trace of acetic acid, also heated to the b o i l i n g point on a steam bath. Reaction was indicated by the fact that a f t e r each addition the orange colour faded to yellow; when the colour no longer faded a f t e r heating for several minutes, i n d i c a t i n g that a s l i g h t excess of the hydrazine was present, water was added to t u r b i d i t y . On standing, a pale yellow s o l i d separated,•which was r e a d i l y Isolated by f i l t r a t i o n . Under the same conditions 3,4-di-O-acetyl- D-xylal did not react, and only unchanged 2,4-dinitrophenyl- hydrazine was i s o l a t e d following d i l u t i o n of the mixture with water. Thin layer chromatography of the yellow s o l i d obtained from t h i s reaction revealed the presence of two major com- ponents, together with traces of others. The two major pro- ducts of the reaction, the fas t e r moving of which w i l l be - 126 - designated Fraction X and the slower, Fraction Y, moved very c l o s e l y together on s i l i c a gel using a variety of developing solvents. The absence of colourless impurities i n the mixture was demonstrated on spraying t h i n layer chromatograms with a s u l p h u r i c - n i t r i c acid reagent and heating. From an amount of 1.2 g of the main f r a c t i o n from the oxo reaction, 0.4 g of 2 , 4-dinitrophenylhydrazones, consisting e s s e n t i a l l y of Fractions X and Y, was i s o l a t e d . On the assumption that X and Y were derivatives of isomeric di-Q - a c e t y l-a nhydrpdeoxy-a ldehydo-hexoses (133) and (134) , i and that d e r i v a t i s a t i o n was roughly quantitative, the weight of hydrazones i s o l a t e d therefore indicated the presence of approximately 15$ of aldehydes i n the mixture. The n.m.r. spectrum of the product obtained by reaction with 2 , 4-dinitrophenylhydrazine confirmed i t s composition . as a mixture of two 2 , 4-dinitrophenylhydrazones of acety- l 6 l lated anhydrodeoxy-aldehydo-hexoses. The readily-assignable low f i e l d signals due to -NH and protons of the dinitrophenyl group showed i d e n t i c a l chemical s h i f t s for both isomers ex- cept f o r the aromatic C-6 proton, which has a large s p l i t - t i n g (J = 9-10 c/s) by coupling with the ortho-hydrogen. A p a i r of overlapping doublets (J = 9 c/s) at $ = 7.87 and 7«95 ppm demonstrated the presence of a mixture of isomers, t h i s also being shown by 3 acetate signals at high f i e l d . The doublet at S =7.85 ppm (J =6 c/s) was assignable to the -N=C-H group and confirmed that the - 127/- 161 derivatives were of aldehydes rather than ketones/. ( a) Separation of Fractions X and Y When the mixture of 2,4-dinitrophenylhydrazones was heated with a small volume of ethanol, p a r t i a l solution of the mixture occurred, leaving a granular yellow s o l i d which was not r e a d i l y soluble i n ethanol. Upon i s o l a t i o n and examination by thin layer chromatography, the insoluble portion showed only one zone corresponding to the slower- moving of the two components (Fraction Y) present i n the mixture. The material thus i s o l a t e d c r y s t a l l i s e d r e a d i l y from chloroform-hexane as f i n e yellow needles, m.p. 225- 226°, [<?<Gn - 6 0 ° , and gave an elemental analysis i n agree- ment with that required for a 2,4-dinitrophenylhydrazone of a 4 ,5-di-0-acetyl-2,6-anhydro-3-deoxy-aldehydo-hexose ((133) or (134)). The n.m.r. spectrum of t h i s f r a c t i o n d i f f e r e d c h i e f l y from that of the mixture i n showing two acetate peaks of equal i n t e n s i t y (§ =2 .13 and 2.17 ppm); a doublet (J= 9 c/s) at £ = 7 . 8 5 ppm, assignable to the C-6 proton of the 2,4-dinitrophenyl group, confirmed the homogeneity of the product. Thus, f r a c t i o n a l solution of the mixture afforded a convenient procedure for the separa- ti o n of Fraction Y from Fraction X i n appreciable amounts, thereby f a c i l i t a t i n g i t s further examination. Treatment of a larger quantity (6 .8 g) of the main f r a c t i o n (from - 128 - the reaction of (22) with 2 moles of synthesis gas) with 2 , 4-dinitrophenylhydrazone as described previously gave 2.4 g of the mixed hydrazones, from which 0 .9 g of pure Fraction Y was obtained by t r i t u r a t i o n of the mixture with warm ethanol, and r e c r y s t a l l i z a t i o n of the insoluble portion. I s o l a t i o n of the other derivative, Fraction X, from the alcohol-soluble residue, which contained both Fractions X and Y, presented a d i f f i c u l t problem because of the simi- l a r m o b i l i t i e s of the two fractions on chromatography. It was possible to i s o l a t e a small amount of the f a s t e r - moving component, X, s u f f i c i e n t for characterisation, by chromatography on thick plates of s i l i c a g el using the technique of multiple development, with chloroform as de- veloping solvent. The derivative i s o l a t e d i n t h i s way had m.p. 1 3 2 ° , - 1 6 ° , and also analysed s a t i s f a c t o r i l y for C-̂ gH-̂ gN̂ Ô . A quantitative determination of the amounts of the two isomeric 2 , 4-dinitrophenylhydrazones i n the mix- ture, by chromatography on s i l i c a gel i n the same way, showed that X and Y were present i n the approximate r a t i o of 1 :2. (b) I d e n t i f i c a t i o n of Fraption Y An adequate amount of Fraction Y being.available, i t was possible to convert t h i s 2 , 4-dinitrophenylhydrazone to the free aldehydo-hexose, which was then i d e n t i f i e d by deacetylation and reduction to the corresponding anhydro- deoxyhexitol of known structure. Conversion of Fraction Y r " - 129' to the free aldehyde was effected by e q u i l i b r a t i o n of the hydrazone with an excess of pyruvic acid, following a proced- 162 applied 'by Mattox and Kendall to the hydrolysis of the 2 , 4-dinltrophenylhydrazone of a s t e r o i d a l ketone. The 2 , 4-dinitrophenylhydrazone of pyruvic acid so formed was re a d i l y removed from the reaction mixture by f i l t r a t i o n , and by extraction with sodium hydrogen carbonate solution. A syrupy product was i s o l a t e d from t h i s reaction which was not completely characterised J however, the fact that i t 159 was an aldehydo-compound was shown by the presence of a sharp signal at S=9*35 ppm i n i t s n.m.r. spectrum, measured i n deuteriochloroform solution. It i s thought that the product probably existed as a mixture of free and hydrated aldehyde, as the i n t e n s i t y of t h i s low f i e l d signal r e l a t i v e to the remainder of the spectrum was only about half that required for a 4 ,5-di-0^abetyl-2,6-anhydro-3-deoxy- aldehydo-hexose ((133) or ( 1 3 4 ) ) . Additional evidence f o r th i s was the presence i n the infrared spectrum of bands assignable to both aldehyde (C=0 stretching at 1700 cm"1 i n addition to acetate absorption at 1740 cm"*'") and hydroxyl groups. That the l a t t e r were not a re s u l t of deacetylation during and subsequent to the exchange reaction with pyruvic acid was shown by the fact that when a portion of the syrupy product was reacted with 2 , 4-dinitrophenylhydrazine, the o r i - g i n a l c r y s t a l l i n e d e r i v a t i v e , Fraction Y, was obtained. Wolfrom 163,164 and co-workers have observed the ready.formation, by 0 - acetyl-aldehydo-sugars, of c r y s t a l l i n e hydrates and alcoholates which have been shown to be aldehydrol (135) and hemiacetal (136) derivatives; t h e i r formation accounts f o r the mutaro- M H yO M H O - C - O H < H Q H Y R Q H > R O - C - O H 6 9 (135) (136) t a t i o n of aldehydo-sugars i n aqueous and alc o h o l i c solutions. The aldehydic product -thus i s o l a t e d was r e a d i l y i d e n t i f i e d as 4 ,5-di-0-acetyl-2,6-anhydro-3-deoxy-aldehydo- D-lyxo-hexose (133)• A portion was reacted with sodium borohydride i n aqueous methanol, whereby the aldehyde group was reduced to hydroxymethyl,. and simultaneous hydrolysis of the a c e t y l groups was effected i n the basic medium. The product Isolated from t h i s reaction was subjected to chroma- tography on paper alongside control spots of the anhydro- deoxyhexitols (44) and (45) , . (Section A), and thereby iden- t i f i e d with the faster-moving isomer, 2,6-anhydro-3-deoxy-D- arabino-hexitol (44) . The aldehydic precursor there^qre: had the D-lyxo-configuration (.133), and Fraction Y was 4 ,5-di-O- acetyl-2,6-anhydro-3-deoxy-aldehydo-D-lyxo-hexose 2,4-dinitro- phenylhydrazone. It followed that Fraction X must be the corresponding derivative of 4 ,5-di -0-acetyl-2,6-anhydro -3-deoxy a 1dehydo-D-xy1o-hexose (134), as the other major component - 131 - ( 4 5 ) , previously Isolated from the hydroxymethylation of 3 , 4-di-O-acetyl-D^xylal, had the L-xylo-configuration. Thus C H P H F r a c t i o n Y ( y ' i IHO-C-H 9 H - C - O H 70 D-iyxo- .( l33) D-arabino-(44) F r a c t i o n X D-xylo- (134) HO-^-HH . H - C - O H 71 H P H L-xylo- (45) It i s of i n t e r e s t that, whereas the anhydrodeoxy- hexitols (44) and ( 4 5 ) , previously i s o l a t e d from the hydroxy- methylation of 3 , 4-di-O-^cetyl-D-xy.lal were present i n nearly equal proportions, the amount of the D-lyxo-hexose (133) i s o l a t e d (as Fraction Y) from the intermediate stage of the reaction was considerably greater than the amount of i t s isomer ( 134) . I t has been shown that the rate of hydroformylatlon of o l e f i n s depends on the a c c e s s i b i l i t y of the double bond 1^, and i t i s reasonable to suppose that a s i m i l a r effect w i l l operate i n the hydrogenation stage, which i s also considered to proceed via an intermediate 20 Tt-complex with cobalt hydrotricarbonyl .. It i s suggested therefore, that the carbonyl group of (13*0, being i n equa- t o r i a l orientation to the r i n g , i s more accessible for com- plex formation, and (134) i s more rapidly removed from the reaction mixture by hydrogenation that i s (133)> In which the formyl group i s presumably a x i a l . i . 1 (45) (44) ( i i i ) Aldehydo-hexoses by Oxidation of Hexitols The attempted hydroformylation experiments described above c l e a r l y showed that the oxo reaction of 3 , 4 - d i - 0 - a c e t y l - D-xylal was not a s a t i s f a c t o r y source of aldehydo-hexoses (133) and (134) , as these were apparently too r e a d i l y reduced to alcohols and did not accumulate i n the reaction mixture. Some experiments were therefore carried out with a view to preparing (133) and (134) by oxidation of the more accessible ii - mixture of dl-O-acetyl-anhydrodeoxyhiexitols (74) and (75) , v r e s u l t i n g from the hydroxymethylation of (22) as described 133 - In Section A. Few methods are available f o r the oxidation of alcohols which stop short at the aldehyde stage? two recently described, and somewhat sim i l a r procedures which were investigated are those of Kornblum and co-workers and of P f i t z e r and M o f f a t t 1 6 6 . 165 Kornblum and co-workers have obtained good y i e l d s of aldehydes by the oxidation of a variety of benzylic and straight chain a l i p h a t i c tosylates, on heating with a mix- ture of sodium hydrogen carbonate and dimethylsulphoxide. The l i m i t a t i o n s of the oxidation, which I n i t i a l l y involves displacement of the tosyloxy group by dimethylsulphoxide, with formation of an Intermediate dimethylsulphoxonium s a l t were shown by the non-reaction of neopentyl tosylate An attempt to oxidise the terminal tosyloxy group of a carbohydrate, l , 2 - 0_-isopropylidene - 6-G-p-tolysulphonyl-D- glucose, did not give the anticipated aldehydo- compound 168 : but the 5>6-carbonate The mixture of d i - 0 - a c e t y l - l , 5 - anhydro-4-deoxy-hexitols (74) and (75) comprising the hydroformylation product from (22) (Section A) was tosylated with p-toluenesulphonyl chloride i n pyridine to give the previously described crude mixture of isomeric 6-G-p- tolylsulphonyl derivatives (76) and (77) (equation 5 2 ) . This was then reacted for 5 minutes at 1 5 0 ° , i n an atmos- phere of nitrogen, with a mixture of sodium hydrogen car- bonate and dimethylsulphoxide, according to the procedure of Kornblum and co-workers. Prom th i s reaction a dark coloured syrup was i s o l a t e d whose inf r a r e d spectrum was v i r t u a l l y unchanged from that of the mixture of 6-0-p- tolylsulphonyl derivatives (76) and ( 7 7 ) . Results of a more po s i t i v e nature were obtained by 166 the procedure of P f i t z e r and Moffatt which also involves the reaction of dimethylsulphoxide. These workers found that complex alcohols (nucleoside derivatives, steroids) were ra p i d l y and s e l e c t i v e l y oxidised to the corresponding aldehyde (or ketone) on treatment with N,N'-dicyclohexyl- carbodlimide (DCC) and dimethylsulphoxide at room tempera- ture, i n the presence of c e r t a i n acids. A solution was prepared of the mixture of (74) and (75) (from the hydroxy- methylation of 3 , 4-di-O-acetyl-D-xylal) i n dimethylsulphoxide containing a trace of anhydrous phosphoric acid and an ex- cess of DCC. After standing f o r one day at room temperature, during which time a c r y s t a l l i n e p r e c i p i t a t e of N,N'-dicyclo- hexylurea formed and dimethylsulphide was evolved, the mix- ture was f i l t e r e d , d i luted with ethanol and treated with a hot ethanolic solution of 2 , 4-dinitrophenylhydrazine, as described previously. On d i l u t i n g with water a yellow s o l i d p r e c i p i t a t e d . This was c o l l e c t e d by f i l t r a t i o n and separated from co-precipitated N,N 1-dicyclohexylurea by solution i n a small volume of chloroform, i n which the urea was Insoluble. Examination of the yellow product by th i n layer chromatography - 1-35 - revealed the presence of two major components, which were is o l a t e d "by chromatography on thick plates of s i l i c a g e l , using multiple development with chloroform, and i d e n t i f i e d with Fractions X and Y, the 2 , 4-dinitrophenylhydrazones of the D-xylo- and D-lyxo- Isomers of 4 , 5 - d i - 0 - a c e t y l - 2 , 6 - anhydro-3-deoxy-aldehydo-hexose respectively, by melting point and mixture melting point, and th i n layer chromato- graphy. The amount of mixed 2 , 4-dinitrophenylhydrazones ob- tained by t h i s reaction was equivalent to the oxidation of approximately 35$ of the di-O-acetyl-anhydrodeoxyhexitols (74) and (75)$ and comprised about 60$ of the D-lyxo- (Y) and about 30$ of the D-xylo- derivative (X). That the aldehydo-hexoses thus Isolated were products of the o x i - dation reaction was confirmed by running separate control experiments under the same conditions, i n one of which the hydroxymethylation product, and i n the other the carbodiimide, were omitted. Neither of these gave derivatives on subse- quent treatment"with 2 , 4-dinitrophenylhydrazine. - 136 - EXPERIMENTAL General Considerations High pressure reactions were carr i e d out using an Aminco 2 ^ / i 6 " o.d. Micro Series reaction vessel of man- ganese s t e e l (Americal Instrument Co. Inc., S i l v e r Spring, Md.). Infrared (I.R.) spectra were measured on Perkin- Elmer Model 21 and Model .137 (Sodium Chloride) spectro- photometers. Nuclear magnetic resonance (n.m.r.) spectra were recorded at 60 Mo/s on a Varian A 6 0 spectrometerJ resonance positions are recorded i n ppm from tetramethyl- silane as reference, set at zero (external with DgO, i n - te r n a l with other solvents). Double resonance spectra were measured on a Varian D. P. spectrometer, also at 60 Mc/s, using a Heteronuclear Decoupler (Nuclear Magnetic Re- sonance S p e c i a l t i e s ) ; modules operating at ca_ 9 .2 Mc/s were used to provide the deuterium frequency. Gas-liquid p a r t i t i o n chromatography (GLPC) separations were effected using an Aerograph "Autoprep" Model A-700 (Wilkens Instru- ment Co. Inc.), employing a column ( 1 0 ' x 1 / ^ " ) °f 2 0 $ S i l i c o n e GE-SF-Q6 on f i r e b r i c k , at a temperature of 1 8 0 ° , with helium as the c a r r i e r gas at a flow rate of 40 ml/ minute:. Samples were injected d i r e c t l y onto the column using a 6" needle. Water-saturated 1-butanol containing 5$ ethanol at room temperature was employed as solvent for -137 - paper p a r t i t i o n chromatography* R p values recorded are with reference to t h i s solvent system. Polyols were detectedSwIth 73 sodium periodate-Schiff reagent . Thin layer chromatography (TLC) was on plates of S i l i c a Gel G (acc. to Stahl), and zones were located by spraying with the general purpose re- agent of sulphuric acid containing 5$ fuming n i t r i c acid, and heating at 1 3 0 ° . Melting points were determined on a Kofler block and are uncorrected. Analyses were performed i n the laboratories of Dr. A. Bernhardt, Mulheim (Ruhr), West Germany, and by the Microanalytical Laboratory, University of B r i t i s h Columbia. - 1 3 8 - Experimental Section A 3,4-Di-O-acetyl-D-xylal (22) In a 500 ml, 3-necked f l a s k equipped with an e f f i c i e n t s t i r r e r and a thermometer, acetic anhydride (200 ml) was cooled to 0° and 70$ perchloric acid (1.2 ml) was added dropwlse. The solution was then warmed to room temperature and D-xylose (50 g) was added i n portions during the course of 1 hour to the s t i r r e d mixture at such a rate that the temperature did not r i s e above 30° . Red phosphorus (15 g) was added a f t e r cooling the reaction mixture to below 20° i n an ice-water bath, followed by the dropwlse addition of bromine (30 ml) and then of water (15 ml) to the con- tinuously s t i r r e d mixture with control of temperature at or below 20°. After standing at room temperature for 3 hours the reaction mixture was f i l t e r e d , and the f i l t e r paper was washed with a l i t t l e g l a c i a l acetic a c i d . The deep yellow f i l t r a t e , containing 2 , 3,4-trl-Q-acetyl-cx-D- xylopyranosyl bromide, was immediately added to a reduction mixture, previously prepared as follows. A solution of sodium acetate trlhydrate (200 g) i n water (290 ml) and g l a c i a l acetic acid (200 ml) was cooled to -10°, and zinc dust (110 g) and cupric sulphate pentahydrate (11 g) dissolved i n water (40 ml) were added. When the blue colour had d i s - appeared, the above solution of 2 , 3,4-trl-O-acetyl-oc-D- - 139 - xylopyranosyl bromide was gradually added over a period of 1 hour to the mixture, which was maintained at -10° by cooling i n an acetone-solid COg bath, and which was vigor- ously s t i r r e d by an e f f i c i e n t Hirschberg-type s t i r r e r . S t i r - r i n g was continued f o r a further 3 hours with the temperature at - 1 0 ° , C e l i t e was added, and the mixture was f i l t e r e d through a layer of C e l i t e into a suction f l a s k containing ca 500 g of crushed i c e , fragments of s o l i d carbon dioxide being added to the mixture during f i l t r a t i o n to prevent undue r i s e i n temperature. After washing the f i l t e r with cold 50$ aqueous acetic acid (ca_ 150 ml), the f i l t r a t e was extracted with f i v e 100 ml portions of cold chloroform. The combined chloroform extracts were washed with f i v e 100 ml portions of i c e - c o l d water, cold aqueous sodium carbonate u n t i l free of acid, three 100 ml portions of cold water, then dried over anhydrous calcium chloride, f i l t e r e d and evaporated under reduced pressure to a syrup, which was. immediately d i s t i l l e d under vacuum. The f r a c t i o n b.p. 1 1 5 - 1 2 0 ° / 1 . 4 mm • 22 (26 g) c r y s t a l l i s e d overnight i n the r e f r i g e r a t o r , Lpdp -312° (e, 2 .3 i n chloroform). One spot by TLC (benzene- methanol, 96^4 v/v) > I.R., strong band at 1640 cm~"L (C=G stret c h i n g ) . Dicobalt Octacarbonyl A s l u r r y of cobalt ( i i ) carbonate (15 g) i n anhydrous benzene (60 ml), contained i n the Pyrex glass l i n e r of a - 140 - high pressure reaction vessel was shaken with carbon mono- xide (1600 psi) and hydrogen ( l 6 0 0 psi) at 180° f o r 2 hours. After cooling to room temperature the unreacted gases (combined pressure about 2400 psi) were vented and the dark brown solution was f i l t e r e d to remove unreacted cobalt (II) carbonate. The f i l t r a t e was diluted with an equal volume of petroleum ether (b.p. 30-60°), and on standing at -10° i n a well stoppered f l a s k , orange cr y s t a l s of dicobalt octa- carbonyl were precipitated, y i e l d about 10 g. The c r y s t a l l i n e o product could be stored under the mother liquors at -10 for several weeks without undue decomposition. Hydroxymethylation of 3,4-di-Q_-acetyl-D-xyla 1 Typical experimental conditions f o r the absorption of 3 moles of synthesis gas were as follows. To a solution of 3*4-di-0-acetyl-D-xylal (5 .6 g) i n anhydrous benzene (25 ml), contained i n the glass l i n e r of a high pressure reaction vessel, was added dicobalt octacarbonyl (1 .5 g)• The stop- pered l i n e r was then inserted Into the autoclave, which was flushed with carbon monoxide. Carbon monoxide was then added to a pressure of 500 p s i , followed by hydrogen to a t o t a l pressure of 3000 p s i , and the reactants were heated, with rocking, at 125-13©° for about 2 hours. After cooling to room temperature overnight, unreacted synthesis gas pressure was released and the dark coloured solution was transferred - .141 - to a short (8 x 8 cm dlam.) column of F l o r i s i l , previously prepared as a s l u r r y i n anhydrous benzene. E l u t i o n with petroleum ether (b.p. 30-60°) was continued u n t i l catalyst was completely removed and the eluate was colourless, and the reaction produet was then eluted with benzene-ethanol (10*1, v/v). Evaporation of solvent under reduced pressure gave a syrup:;- '($''*&/. g));c.o$s£|ping p r i n c i p a l l y of a mixture of 2 , 3-di - 0-acetyl-l,5-anhydro-4-deoxy-hexitols. To the syrupy product (4.4 g) dissolved i n anhydrous methanol (50 ml) was added, with cooling, a solution of sodium ( 0 . 2 g) i n anhydrous methanol (5© ml), and the mixture o was set aside at about 5 f o r 1 day. Fragments of s o l i d carbon dioxide were then added u n t i l the solution was neutral to pH ind i c a t o r paper, and solvent was removed under reduced pressure. The r e s u l t i n g s o l i d residue was dissolved In water (50 ml) and sodium ions were removed by passage of the solution through a column of Amberlite IR-120 (H +) cation exchange r e s i n , which was then washed with d i s t i l l e d water. The combined effluent and washings ( t o t a l volume about 25© ml) were then reduced to a syrup (2.75 g) by evaporation under reduced pressure at about 45°. The pro- -1 duct, which exhibited a strong band at around 3400 cm~ (0-H stretching) i n the i n f r a r e d , showed two main components on paper chromatography. A portion (0.40 g) of the deacety- lated mixture, dissolved In a small volume of methanol, was - 142 - applied equally to sijjp; sheets (57 x 46 cm) of Whatman No. 1 f i l t e r paper, prepared for descending chromatography. After e q u i l i b r a t i o n and development f o r 40 hours, the two zones were located by spraying tests s t r i p s , cut from each sheet, with aqueous sodium periodate-Schiff reagent, and separately extracted with hot aqueous methanol ( i : i , v/v) to afford Fraction I (0.15 g), jR p 0 . 4 7 , and Fraction II (0.18 g), R p 0 . 4 1 . " Characterisation of Fractions I and II Fraction I (1,5-anhydro-4-deoxy-D-arablno-hexitol (44)) Fraction I, R p 0 . 4 7 , i s o l a t e d by paper chromatography, was dissolved i n methanol, decolourised by f i l t r a t i o n through a layer of Gelite-Darco 6 0 , and c r y s t a l l i s e d by the addition of isopropyl ether to in c i p i e n t t u r b i d i t y and cc'oling i n the r e f r i g e r a t o r . R e c r y s t a l l i s a t i o n from the same solvents r ~i2® o gave a product, m.p. 1 0 2 ' , LPClp -13 (e, '6 .3 i n water) . N.m.r. signals (DgO): multiplet 3 .2 - 4 . 2 ( 7 ) ; multiplet 1.35 - 2.2 ppm ( 2 ) . Calc. for C gH 0^: C, 48 .64 ; H, 8 . 1 6 . Found: C, 4 8 . 3 5 ; H, 8 . 1 0 . 1,5 - Anhydro-4-deoxy-2,3,6-t r 1-0-p_-nit r obenz oy 1-D- arabino-hexitol Fraction I (31 mg) and p-nitrobenzoyl chloride (freshly d i s t i l l e d , 330 mg) were dissolved i n anhydrous pyridine - 143 - (1 ml) and the solution was heated at 80-90° f o r 1 hour. After cooling, the mixture was s t i r r e d with a saturated aqueous solution of sodium hydrogen carbonate (5 ml) for 30 minutes. The l i g h t coloured s o l i d which separated was extracted with chloroform, and the extract was washed with saturated sodium hydrogen carbonate solution and with water, and dried over magnesium sulphate. Removal of solvent gave a syrup which c r y s t a l l i z e d from ethyl acetate-petroleum ether (b.p. 30-60°) ; m.p. 2 1 5 ° ; ~ 5 ° ° ( c * 1 , 0 l n c h l o r ° - form). Calc. f o r 0 H 0 N : C, 5 4 . 6 0 ; H, 3.6©. Found: 27 21 13 3 C, 5 4 . 5 6 ; H, 3 . 8 6 . Fraction II (1 ,5-anhydro -4-deoxy-L-xylo-hexitol (45)) The component, R^ 0 . 4 l , was obtained as a syrup, a f t e r f i l t r a t i o n through charcoal, which could not be c r y s t a l l i s e d . P u r i f i c a t i o n by formation of a c r y s t a l l i n e t r i - O - a e e t y l d e r i v a t i v e , as described below, followed by deacetylation with methanolic sodium methoxide, afforded r n 2 0 o a colourless syrup; LP*Jj) -44 (c, 6.4 i n water). N.m.r. signals (D 2 0 ): multiplet 2 .9 - 4.15 ( 7 )> pair of quartets 1.75 - 2.2 ( 1 ) ; multiplet 1.15 - 2.2 ppm ( l ) . Gale, f o r C 6 H 1 2 ° 4 : C> 4 8 - 6 4 j H> 8 ' l 6 « Pound: C, 4 8 . 3 6 ; H, 8 . 3 9 . 2 , 3 * 6-Tri - 0-a c ety1-1 ,5-a nhydro -4-deoxy-L-xylo-hexitol Fraction II (70 mg) i n pyridine ( l ml) and acetic - 144 - anhydride ( l ml) was heated for 20 minutes on a steam bath with exclusion of moisture. The solution was kept at room temperature overnight and then poured into ice-water (about 50 ml). After 1 hour the aqueous solution was extracted with chloroform, which was successively washed with 5$ aqueous potassium hydrogen sulphate solution, saturated sodium hydrogen carbonate solution and water, and dried over magnesium sulphate. Removal of solvent gave a syrup (110 mg) which soon c r y s t a l l i s e d . Recrystallisjation from ether-petroleum ether (b.p. 30-60°) afforded the pure t r i - o 22 acetate; m.p. 80-81 ; [p{] _ 4l° (c, 0 .8 i n chloroform). Calc. for C^H^O : C, 52.55? H, 6 . 6 2 . Found: C, 5 2 . 8 2 ; H, 6 . 5 8 . Consumption of Periodate Ion' Fraction I Absorbance readings at 223 myu. were measured at i n - tervals on a Beckmann Model DU Spectrophotometer (1 cm s i l i c a c e l l s ) of an aqueous solution containing 0.439 x 10~^M of Fraction I and 0.942 x 10 M of sodium periodate (Reading A). Simultaneous readings were made of a solution containing 0.439 x 10""̂ M of Fraction I only (B), and of a solution con- t a i n i n g 0.942 x 10~^M of sodium periodate (C). The following values were obtained: - 145 Time (hrs) (B C)-A<a) (B C ) - A W C 0 Moles periodate/ moleV substrate . 1.5 0.142 0.150 0.32 3.5 0 .207 0.218 0.47 11.5 0.399 0.420 0 .90 22 0 .381 0.401 0 .86 (a) : decrease In absorbance due to consumption of periodate (b) : f r a c t i o n of known amount of periodate consumed Fraction II In a s i m i l a r manner, absorbances of solutions -4 -4 0.453 x 10 M with respect to Fraction II and 0.942 x 10 M with respect to periodate ion were measured, to give the following values: Time (hrs) . (B C)-A( a) , (B C)-A(to) C o Moles periodate/ mole substrate 1.5 0.177 0.192 0.40 3 . 0 0.263 0.284 0.59 10.5 0.423 0.457 0.95 24 0.392 0.424 0.88 - 146 _ Enantiomeric 2-deoxy - 3 - 0 - ( 2-hydroxyethyl)-glycero-tetritols Fraction I, Rp 0.47 (47 mg) was dissolved In an aqu- eous solution (5 ml) containing periodic acid (150 mg, 5C# excess) and immediately transferred to a polarimeter tube which was protected from l i g h t . The o p t i c a l rotation of the solution rapidly assumed a constant value. After 6 hours the solution was removed from the tube, neutralised by the addition of excess barium carbonate, and f i l t e r e d into a solution of sodium borohydride (5© mg) i n water (3 ml). After standing at room temperature f o r 1 "Vg hours the solution was made s l i g h t l y acid to pH indi c a t o r paper by the dropwlse addition of acetic acid, Amberlite IR-120 (H +) r e s i n (5 ml) was then added, and the mixture was s t i r - red f o r 1 hour. F i l t r a t i o n and evaporation under reduced pressure gave a s o l i d residue, which was repeatedly eva- porated with methanol to remove borate ion as the methyl ester. The r e s u l t i n g t r i o l ether (III) (44 mg) was a syrup; L°<]jp - 1 9 ° (c, 2 .0 i n water); n.m.r. signals (D 2 0) : multiplet 3.5 - 3 . 9 , with sharp signal at 3.72 ( 9 ) ; multi- plet (apparent quartet) 3/fji" - 2 .0 ppm ( 2 ) . The t r i o l ether ( i l l ) was characterised as the t r i s - p-nitrobenzoyl derivatives a portion (18 mg) was heated at 90° with p-nitrobenzoyl chloride (160 mg) i n anhydrous pyridine (0 .6 ml) f o r l h o u r . Removal of excess p - n i t r o - - 147 - benzoyl chloride and pyridine by s t i r r i n g with saturated sodium hydrogen carbonate solution, extraction with chloro- form, washing with sodium hydrogen sulphate solution, sodium hydrogen carbonate solution and water, drying over magnesium sulphate and removal of solvent gave a syrup (80 mg) which was c r y s t a l l i s e d from ethyl acetate-petroleum ether (b.p. 22 30-60°) ; m.p. 102-103°; M D -28° (c, 1.1 i n chloroform). Calc. f o r C o 7 H o o 0 N : C, 5 4 . 3 0 ; H, 3 - 9 9 . Found: C, <o 13 3 5 4 . 5 7 ; H, 3 . 9 5 . The melting point of t h i s derivative was undepressed on admixture with the corresponding derivative of the t r i o l ether obtained, by a si m i l a r procedure to the above, from the polyols i s o l a t e d from the hydrogenolysis 82 products of methyl CX-D-glucopyranoside Fraction I I , R„ 0 . 4 l , (3© mg) was converted by perio- date oxidation and sodium borohydride reduction as described above to a t r i o l ether (IV) (27 mg), which had |jx] ̂  +17 (c, 1.7 In water); n.m.r. spectrum i d e n t i c a l with that de- scribed above f o r t r i o l ether (III) from Fraction I. A portion (13 mg) of t r i o l ether (IV) from Fraction II was converted to a tris-p-nitrobenzoyl derivative on heating with p-nitrobenzoyl chloride (0.13 g) i n pyridine (0.5 ml), and i s o l a t e d i n the usual way; the product (55 mg) on cry- s t a l l i s a t i o n from ethyl acetate-petroleum ether (b.p. 3 0 - 60°) had m.p. 102-103°; [p<l^ +26° (c, 1.4 i n chloroform); - 148 - infrared spectrum i d e n t i c a l with that of the t r i s - p - n i t r o - benzoate of t r i o l ether ( I I I ) . Calc. for C 2 7 H 2 3 ° 1 3 H 3 : 0 , 5 4 . 3 0 ; H, 3 . 9 9 . Pound: c , 54.63) H, 4 . 0 9 . 2-Deoxy -3-0-(2-hydroxyethy1)-L-glycero-tetr11ol (49) D-Oalactal (19) 3 , 4 , 6-Tri-O-acetyl-D-galactal (40)(7 g) was dissolved i n a solution (approximately 0.01 N) of sodium methoxide i n anhydrous methanol, and the solution was kept at room temperature f o r 48 hours. Removal of solvent under reduced pressure gave a syrup, from which D-galactal was i s o l a t e d as white c r y s t a l s on extraction with b o i l i n g ethyl acetate. R e c r y s t a l l i S a t i o n from the same solvent gave ( 19)(2 .5 g), m.p. 100-102°. Methyl 2-deoxy-<x-D-galactopyranoside (55) To a solution of D-galactal (2.5 g) i n anhydrous methanol (22.5 ml) was added a 2% (w/v) solution of hydrogen chloride i n methanol (2.5 ml). A portion of the solution was transferred to a 2 dm polarimeter tube and the change i n o p t i c a l r o t a t i o n was observed at i n t e r v a l s ; no further change was observed a f t e r 90 minutes from the addition of hydrogen chloride. After 2 hours the recombined solutions were neutralised by the addition of s i l v e r carbonate, f i l t e r e d and evaporated to a syrup (2.7 g). Addition of a small volume - i 4 9 - of acetone resulted In the separation of c r y s t a l s which were is o l a t e d and r e c r y s t a l l i s e d from ethyl acetate; 1.0 g; m.p. 1 1 2 -114°. Methyl 3,6-anhydro-2-deoxy-oc-D-galaetopyranoside (57) To an ic e - c o l d , s t i r r e d solution of methyl 2-deoxy- oc-D-galactopyranoslde ( 1 . 0 g) In anhydrous pyridine, a cooled solution of p-toluenesulphonyl chloride (l.O g) i n anhydrous pyridine (4 ml) was added over a period of 50 minutes, and the mixture was set aside at 0° f o r 20 hours. Water (0 .5 ml) was added to the s t i r r e d solution at 0 ° , which a f t e r 30 minutes was then poured into ice-water (100 ml). The mixture was extracted with chloroform, and the combined extracts were washed with potassium hydrogen s u l - phate solution, sodium hydrogen carbonate solution and water, dried over magnesium sulphate, f i l t e r e d and evaporated to a syrup (56) (1.46 g). This was dissolved i n ethanol (53 ml); measurement of the o p t i c a l rotation of t h i s solution gave a value of Qcxfljp + 93° (c, 2.6 i n ethanol) for the 6 - 0_-p-tolylsulphonyl derivative (56). To the recombined ethanolic solution was added IN sodium hydroxide solution ( 4 . 3 ml), and the solution was heated at 60° f o r 1 hour, neutralised with s o l i d carbon dioxide and evaporated to dryness. The product was repeatedly extracted with b o i l i n g acetone and a f t e r removal of solvent under reduced pressure - 150 - the residue was redissolved In ethyl acetate. F i l t r a t i o n and evaporation of solvent gave a syrup (O.65 g), which d i s t i l l e d at 130° (bath temperature )/0.1 mm as a colourless l i q u i d which soon c r y s t a l l i s e d ; y i e l d 0.47 g of methyl 3 , 6 - anhydro-2-deoxy-oc-D-galactopyranoside (57)* which was re- o 1— -124 o c r y s t a l l i s e d from ether; m.p. 76-77 > L°<J D +94 ( 0 , 5 .7 i n water). 3,6-Anhydro-2-deoxy-D-lyxo-hexose (58) To a solution of the methyl glycoside (57) (0.28 g) i n water ( 6 . 0 ml) was added 2N sulphuric acid solution (0 .6 ml), and the mixture was l e f t stand at room temperature for 110 minutes. Neutralisation with barium carbonate, f i l t r a t i o n and removal of water by freeze-drying gave a colourless syrup, which was redissolved i n water, f i l t e r e d , and again evaporated by freeze-drying. The residue was dissolved i n warm acetone, f i l t e r e d and evaporated under reduced pressure to a clear, colourless syrup. Y i e l d of 3,6-anhydro-2-deoxy-D-lyxo-hexose, 0.26 gj • LPdp + 2 5 ° (c, 5 .1 In water). I.R. spectrum ( l i q u i d f i l m ) : sharp peak at 1715 cm - 1 (aldehyde C=0 stretching). 1,4-Anhydro-5-deoxy-D-arabino-hexitol (59) To a solution of the anhydro-sugar (58) (95 mg) i n water (l.O ml) was added dropwise a solution of sodium - 151 - borohydride (5© mg) In water (1 .0 ml). After 1 hour, Amber- l i t e IR-120 (H +) r e s i n was added i n small portions u n t i l Jiydrogen was no longer evolved; more r e s i n (ca_ 3 ml) was then added and the mixture was s t i r r e d f or 30 minutes. F i l t r a t i o n and evaporation under reduced pressure gave a s o l i d residue which was repeatedly evaporated with methanol to afford 1 ,4-anhydro-5-deoxy-D-arabino-hexitol (59) as a colourless syrup (86 mg); Qoc]|p +21° (c, 1.7 i n ethanol). The anhydrodeoxyhexitol was characterised as the t r i s - p - nitrobenzoyl derivative. 1 , 4-Anhydro - 5-deoxy - 2 , 3 , 6-tri - 0-p-nitrobenzoyl-D- arabino-hexitol A portion (25 mg) of (59) was heated with p - n i t r o - benzoyl chloride (0.25 g) i n pyridine (0 .7 ml) and the pro- duct was i s o l a t e d i n the usual way. The r e s u l t i n g syrup (114 mg) c r y s t a l l i s e d from chloroform-petroleum ether (b.p. 30-60°) ; m.p. 1 5 9 - 1 6 0 ° ; M D -96 (c, 0.7 i n chloroform). Calc. f o r C 2 7 H 2 1°13 N 3 : °> 54.46; H, 3-56. Found: C, 54 .61 ; H, 3 . 8 6 . 2-Deoxy - 3 - 0 - ( 2-hydroxyethyl)-L-glycero-tetrltol (49) 1,4-Anhydro-5-deoxy-D-arabino-hexitol (59)(55 mg) was dissolved i n a solution of periodic acid (96 mg, 1.3 moles) - 152 - In water ( 5 . 0 ml). The solution was transferred to a p o l a r l - meter tube protected from l i g h t and i t s ro t a t i o n was observed at i n t e r v a l s and found to be constant a f t e r 3 "Vg h°urs. After standing i n the dark overnight the solution was neu- t r a l i s e d with barium carbonate and then f i l t e r e d into a solution of sodium borohydride (70 mg) i n water (4 ml). After 2 hours the solution was neutralised with acetic acid, deionised by s t i r r i n g with Amberllte I R - 1 2 0 ( H + ) r e s i n , f i l t e r e d and evaporated under reduced pressure to a residue which was repeatedly evaporated with methanol to aff o r d a syrup (51 mg), +17° (c, 2 .0 i n water), whose n.m.r. spectrum (DgO) was i d e n t i c a l with that described for t r i o l ether ( i l l ) , and with that of t r i o l ether (IV). A portion (30 mg) of (49) was converted to a t r i s - p-nitrobenzoyl derivative i n the usual way. C r y s t a l l i s a t i o n from ethyl acetate-petroleum ether (b.p. 30-60°) gave a r " i 2 4 o \ • product, L P < J D + 2 7 (c, 1.1 i n chloroform), whose melting point of 101-102° was undepressed on admixture with the t r i s - p-nitrobenzoate of the t r i o l ether (IV) derived from Fraction I I . The inf r a r e d spectra of both nitrobenzoates were iden- t i c a l . Calc. f o r C 2 Y H 2 3 0 1 3 N 3 : C, 54.30? H , 3 . 9 9 . Found: C, 5 4 . 6 i ; H , 4 . 0 0 . - 153 - Attempted preparations of 2-deoxy - 3 , 4-di - 0-acetyl- D-xylopyranosyl cyanide (a) By addition of HCN to 3 , 4-di-O-acetyl-D-xylal ( 2 2 ) . Hydrogen cyanide was generated by addition of a saturated solution of sodium cyanide to 50$ sulphuric acid s o l u t i o n 1 0 0 ; the gas was dried by passage through a series of calcium i o chloride tubes surrounded by water at 30-40 , and then con- densed into a 50 ml f l a s k , protected from atmospheric moisture, containing 3 , 4-di - 0_-acetyl-D-xylal (0.52 g) and sodium cya- nide (10-15 mg)-, by passage through a .coil surrounded by an i c e - s a l t mixture. After the addition of about 5 ml of HCN, the f l a s k was sealed by a calcium chloride tube, -the solu- t i o n was s t i r r e d at 0 for 5 hours, and the hydrogen cyanide was then allowed to evaporate overnight. The I,R. spectrum of the r e s i d u a l syrup was unchanged from that of 3 , 4 - d i - 0 - acetyl-D-xyla 1 . (b) By reaction of 3 , 4-di - 0-acetyl-D-xylopyranosyl chloride with Hg(CN) 2. 3 ,4-Di - 0-acetyl-D-xyl&pyranosyl chloride ( 6 7 ) : a cooled solution of 3 , 4-di-O-acetyl-D-xylal (2 .5 g) In an- hydrous benzene (20 ml) was saturated with dry hydrogen chloride. After standing f o r 1 hour, the solvent was re- moved under reduced pressure at 30° to give a colourless o i l . - 154 - I,R|. spectrum ( l i q u i d film)*, disappearance of absorption at 1640 cm"1 (C=C stretching). To a solution of (67)(1 .5 g) In anhydrous nitromethane (5 ml) was added mercuric cyanide (1.7 g) * and the mixture was s t i r r e d at room temperature for 24 hours with exclusion of moisture. The dark coloured solution was reduced i n volume under vacuum, methanol (15 ml) was added and the mix- ture was poured into water (40 ml) and ice (30 g) containing sodium chloride (2 .5 g). Extraction with chloroform, wash- ing with water, drying over sodium sulphate and evaporation of solvent gave a dark brown syrup, which was chromatographed through a column of neutral alumina using benzene-ether- methanol (70*-30'l, v/v) as developing solvent to afford a nearly colourless syrup. T.L.C. (benzene-methanol, 96"4 v/v) i complex mixture* I.R. spectrum ( l i q u i d f i l m ) ; no absorption i n 2000-2300 cm region (C=N)* N, approxi- mately 2 . 6 $ . In a s i m i l a r experiment, a solution of ( 6 7 ) ( l . l g) i n anhydrous benzene (7 ml) was added dropwise to a re- flux i n g suspension of s i l v e r cyanide (0 .7 g) i n anhydrous ether (15 ml). After 5 hours the mixture was cooled, f i l t e r e d and evaporated to a syrup which consisted of a mixture of unreacted (67) and 3 , 4-di-O-acetyl-D-xylal ( i d e n t i f i e d by I.R. band at 1640 cm - 1, and T . L . C ) . - 155 - 2,3-Dl-0-ac ety 1 - 1 , 5 -a nhydro-4,6-dideoxy-B-arablno- h e x l t o l (70) and 2 , 3-dl - 0-acetyl - 1 , 5-anhydro - 4 , 6 - dideoxy-L- xylo-hexltbl (71) : (a,) By reaction of (67) with methyl magnesium "bromide. To a s t i r r e d solution of methyl magnesium bromide i n dry ether, prepared by the slow addition of methyl bromide gas to a s t i r r e d suspension of magnesium (3 .2 g) i n dry ether (50 ml), under an atmosphere of nitrogen, was added drop- wise over 45 minutes a solution of 3 , 4-di - 0-acetyl - 2-deoxy- D-xylopyranosyl chloride (prepared by the addition of hydrogen chloride to 3 , 4-di-O-acetyl-D-xylal (2 .5 g) as described above) i n dry ether (30 ml). After r e f l u x i n g f o r 5 hours, the solution was cooled and added slowly to a mixture of ice and water (approximately 750 ml) containing acetic acid (5 ml). The aqueous phase was f i l t e r e d through C e l i t e , neutralised with 2N NaOH solution, and evaporated to a white s o l i d . The reaction product was separated from Inorganic material by repeated extraction, f i l t r a t i o n and evaporation with ethanol, and subsequently with acetone, to y i e l d a yellow syrup (0.95 g)• Paper chromatography showed one zone (Rp O.69) on spraying with periodate-Schiff reagent. A portion of the product (0 .33 g) was p u r i f i e d by chromatography on paper to give a mixture (0.22 g) of anhydrodideoxyhexitols (70) and (71) . N.m.r. signals (DO): complex multiplet - 156 . 3.0 - 4.2 (5)" multiplet u p - f i e l d from 2.2 (2), C-CH0-C; pair of overlapping doublets (J = 6 c/s) at 1.14 and 1.17 Ppm ( 3 ) , C-CH3. A portion of the p u r i f i e d mixture of (70) and (71) (100 mg) was acetylated with acetic anhydride ( l ml) and pyridine ( l ml) at room temperature overnight. I s o l a t i o n of the product i n the usual way gave a syrup (110 mg) which was fractionated by GLPC to y i e l d two pure components (72) and (73)* the l a t t e r being i d e n t i f i e d as 2 , 3-dl-0- acety .1 -1,5-anhydro-4 ,6-dideoxy-L-xylo-hexitol by independent synthesis from 1 ,5-anhydro -4-deoxy-L-xylo-hexitol (45) as described below. 2,3-Di-0-a c ety1-1,5-anhydro-4,6-dideoxy-D-arabino- h e x i t o l (72)> r e l a t i v e retention time 1.10" colourless l i q u i d ; 11°^^ " 8 2 ° (°* 1 , 0 l n chloroform). Calc. for c 1 0 H l 6 ° 5 : °V 5 5 . 5 4 ; H, 7.46. Pound: C, 55.61s H, 7 . 5 5 . N.m.r. signals (CCl^) : multiplets at 4.75 - 5.0 (2), 4.4 - 4.65 (1), 3.35 - 3 .9 ( 3 ) ; two sharp signals at 2.02 (3) and 2 .06 ( 3 ) , -0C0CH ; multiplet 1 .5 - 1 .85 (2), C-CH0-C; doublet 3 d (J = 6 c/s) 1.13 ppm ( 3 ) , C-CH_. ~> i 2,3-Di-0-a c ety1-1,5-a nhydro-4,6-dideoxy-L-xylo-hexitol (73) : r e l a t i v e retention time 1.00; colourless l i q u i d ; r ~i 24 L ° y D -21°(c , 0 .9 i n chloroform). Calc. f o r C^^gO^: C, - 1 5 7 - 5 5 . 5 4 J H, 7.46. Found: C, 5 5 . 4 7 ; H, 7.42. N.m.r. signals (GCl^): multiplets at 4 . 6 5 - 5.0.(2), 3-8 - 4 . 2 (l) and.3 . 0 - 3 . 6 5 ( 3 ) : sharp signal at 1 . 9 7 ( 6 ) , -OCOCH^, superimposed on multiplet ( 2 ) , C-CRg-C: doublet (J = 6 c/s) 1.20 ppm ( 3 ) , C - C H . ( 7 3 ) from 1 ,5-anhydro -4-deoxy-L-xylo-hexitol ( 4 5 ) To ah ic e - c o l d solution of ( 4 5 ) ( 7 0 mg) i n anhydrous pyridine ( 0 . 5 ml) was added over 2 0 minutes a solution of p-toluenesulphonyl chloride ( l © 5 mg, 1 . 1 moles) i n anhydrous pyridine ( 1 . 0 ml). After 1 8 hours at 5 ° , acetic anhydride ( 1 . 0 ml) was added and the mixture was stood f o r a further 1 2 hours at room temperature. I s o l a t i o n of the product i n the usual manner gave a syrup (140 mg) which contained approximately 80$ of 2 , 3-dl - 0-acetyl - . 1 , 5-anhydro - 4-deoxy— 6 - 0 - p - t o l y l s u l p h o n y l - L - x y l o - h e x i t o l ( 7 7 ) , as judged by the amount of sodium p-toluenesulphonate li b e r a t e d therefrom,- The crude product i n acetone ( 2 ml) was heated with sodium iodide (140 mg) i n a sealed tube at 1 0 0 ° f o r 3 hours. After cooling, sodium p-toluenesulphonate ( 5 8 mg) was removed by f i l t r a t i o n and the solution was evaporated under reduced pressure to a s o l i d residue Extraction with r e f l u x i n g ether and evaporation of solvent gave a syrup ( 1 2 5 mg)• The crude 6-deoxy - 6-iodo-derivative ( 7 9 ) was dissolved i n methanol (.10 ml) containing 5N sodium hydroxide solution ( 0 . 5 ml), and - 158 - shaken with hydrogen at atmospheric pressure and temperature i n the presence of Raney n i c k e l (15 mg) f o r 30 minutes. F i l t r a t i o n , n e u t r a l i s a t i o n with s o l i d carbon dioxide and evaporation under reduced pressure gave a white s o l i d , which was acetylated by heating at 100° for 3 hours with acetic anhydride (2 ml) and anhydrous sodium acetate (0.4 g). Iso- l a t i o n of the product i n the usual way gave 2 , 3-dl - 0-acetyl- l ,5-anhydro - r4 ,6-dideoxy-L-xylo-hexitol (73) as a syrup (50 mg), which showed one zone on GLPC whose retention time was i d e n t i c a l with the faster-moving component of the mixture of acetates prepared as described above. The infrared spec- trum of a portion of the product thus p u r i f i e d by GLPC was i d e n t i c a l with that of the l a t t e r zone, which was thus i d e n t i f i e d as the L-xylo isomer (73). (b) From (74) and (75) obtained by reaction of 3,4-di-O- acetyl-D-xylal with carbon monoxide and hydrogen 3,4-Di-O-acetyl-D-xylal (5.6 g) i n anhydrous benzene (25 ml) was reacted with carbon monoxide (1100 psi) and hydrogen (2200 psi) at 130° for 3 hours i n the presence of dicobalt octacarbonyl (1.5 g)'» and the product was separated from catalyst as previously described to give a syrup (6.1 g) consisting c h i e f l y of two isomeric 2,3-di -0-acety1-1,|-anhydro- 4-deoxy-hexitols (74) and (75). A portion of the product (1.5 g) was deacetylated with methanollc sodium methoxide. -, m - I s o l a t i o n i n the usual way and f r a c t i o n a t i o n by preparative paper chromatography gave 1,5 -anhydro-4-deoxy-D-arabino- h e x i t o l (44) and 1,5-anhydro-4-deoxy-L-xylo-hexitol (45), i d e n t i c a l with the fractions obtained i n a previous ex- periment . To a further portion of the oxo product (1.5 g) i n anhydrous pyridine (6 ml) was added p-toluenesulphonyl chloride (1 .8 g). After standing at room temperature for 18 hours the product was i s o l a t e d i n the usual manner to y i e l d a syrup (1.6 g) whose n.m.r. spectrum indicated approxi- mately 75$ of 2,3-di-0-acetyl-1,5-anhydro-4-deoxy-6-0-t(p- toly l s u l p h o n y l ) - h e x i t o l s ((76) +(77)) to be present. The product (1.5 g) i n acetone (12 ml) was heated with sodium iodide (1.5 g) at 100° for 3 hours i n a sealed tube. After cooling, sodium p-toluenesulphonate (0.52 g) was removed by f i l t r a t i o n , the f i l t r a t e was evaporated to dryness and repeatedly extracted with b o i l i n g ether. The crude mixture of 2 , 3-di - 0-acetyl-l , 5-anhydro-4,6-dideoxy - 6-iodo-hexitols ((78) +• (79)) (1.0 g) was dissolved i n methanol (40 ml) con- taining 5N sodium hydroxide solution (2 ml), and shaken with hydrogen at atmospheric pressure and temperature i n the presence of Raney n i c k e l (100 mg) for 30 minutes, when ab- sorption of hydrogen (46 ml) was complete. After standing at room temperature f o r 2 hours the solution was f i l t e r e d , neutralised with s o l i d carbon dioxide and evaporated to a white s o l i d . Repeated extraction, f i l t r a t i o n and evaporation with - 160; - i acetone gave a pale yellow syrup, which was further p u r i f i e d by preparative paper chromatography to y i e l d a colourless syrup (0.20 g)J one zone on paper chromatography of i d e n t i c a l R p value (O.69) to the mixture ((70) + (71)) previously ob- tained by reaction of (67) with methyl magnesium bromide. N.m.r. spectrum (D 20): e s s e n t i a l l y i d e n t i c a l with that described above f o r the mixture of anhydrodideoxyhexitols (70) and (71). A portion of the p u r i f i e d product (110 mg) was acety- lated with acetic anhydride-sodium acetate to y i e l d a syrup (140 mg) which was fractionated by GLPC into (74) and (75), i d e n t i c a l with the two fractio n s i s o l a t e d as described above on the basis of retention times, s p e c i f i c rotations ( Bxj§^ -81° and -23° respectively (c, 1 .2 i n chloroform)), and n.m.r, and infrared spectra. I d e n t i f i c a t i o n of "Polyol Y" from Hydrogenolysis of 80 82 Methyl Oc-D-glucopyranoside ' The tris-p-nitrobenzoate of thi s compound (170 mg), kindly supplied by Dr. P. A. J . Gorin, was debenzoylated by r e f l u x i n g with 0.1N methanollc sodium methoxide (30 ml) for 1 hour. F i l t r a t i o n to remove methyl p-nitrobenzoate, and i s o l a t i o n of the product i n the usual way gave a colour- 22 less syrup (38 mg) i +40° (c, 2.2 i n water), (previously 82 0. J . reported value was +19 ) . N.m.r. signals (p^0) • multi- plet 2 .9 - 4.15 ( 7 ) ; pair of quartets centered around 2 .0 - 161 - (l) J multiplet ca_ 35 c/s wide centered around 1.5 ppm (l) . The spectrum was Identical with that of 1 ,5-anhydro-4- deoxy-L-xylo-hexitol ( 4 5 ) . 2 ,3 ,6-Tr1-0-a c ety1-1 ,5-a nhydro-4-de oxy-D-xylo-hexitol The anhydrodeoxyhexitol obtained on debenzoylation as described above was acetylated at room temperature for .18 hours with acetic anhydride ( l ml) and pyridine ( l ml). I s o l a t i o n of the product gave a syrup which c r y s t a l l i s e d from ether-petroleum ether (b.p. 30-60°)' m.p. 80-82°> r ~i22 o LPdjj + 40 (c, 2 .0 i n chloroform). The i n f r a r e d spectrum of the product was Identical with that of the t r i - O - a c e t y l derivative of the levorotatory L-lsomer ( 4 5 ) . Reaction of 3 , 4-DI-O-acetyl-D-xylal with Carbon Monoxide and Deuterium 3 , 4-Di-O-acetyl-D-xylal (1 .8 g) i n anhydrous benzene (8 ml) was reacted with carbon monoxide (.1000 psi) and deuterium (800 psi) at 130° for 3 ^/g hours i n the presence of dicobalt octacarbonyl (0.4 g). After cooling overnight, unreacted gases were vented, and catalyst was removed by f i l t r a t i o n through a short column of F l o r i s i l as described previously, to afford a syrupi, (1 .8 g) on removal of eluting solvent under reduced pressure. A portion (0 .8 g) of the product was deaeetylated with methanolic sodium methoxide - 162 - and on working up i n the usual way a syrup (O.56 g) was is o l a t e d whose Infrared spectrum showed the presence of deuterium (C-D stretching i n 2200-2300 cm - 1 region). Paper chromatography showed 2 main components of si m i l a r m o b i l i t i e s to (44) and ( 4 5 ) , chromatographed alongside as controls. Fractionation of a portion (0.48 g) of the mix- ture of deuterated hexitols by paper chromatography gave the two components, which were separately p u r i f i e d by further chromatography on paper. The two pure fractions were sub- jected to n.m.r. analysis as described below. 2 1,5-Anhydro-4-deoxy-D-arabino-hexito1-4,6,6-H3 (83) R_ 0 . 46 ; M S 3 - 1 1 ° (c, 3-1 i n water). N.m.r. r JJ signals (D g0): m u l t i p l e t . 3 . 3 - 4 .0? ( 5 ) ; unresolved signal centered at 1.53 ppm ( l ) , C-CBH-C. On removal of coupling between hydrogen and deuterium the l a t t e r one proton signal showed as an unresolved t r i p l e t . 1,5- Anhyojr o - 4-deoxy-L-xylo-hexitol - 4 , 6 , 6-H 2 (cis) (84) R p 0 .40 ; [ ° d p 3 -^5° (c, 2.4 i n water). N.m.r. signals (DgO); multiplet 3 . 0 - 4.15 (5)J unresolved signal centered at I . 9 6 ppm ( l ) , C-CBH-C. On removal of coupling between1 hydrogen. and deuteriumthe. l a t t e r one proton s i g n a l was resolved into a quartet whose spacing -was consistent with a pair of l i n e s separated by 5 . 1 c/s which - 163 - were each s p l i t into two further l i n e s 2 .3 c/s apart. 2 , 3 , 6 - T r i -0 -a c e t y l -1,5 -anhyd.ro -4 -deoxy-L-xylo - 2 h e x i t o l - 4 , 6 , 6 - H 3 (els) The deuterated anhydrodeoxyhexitol (84) (37 mg) was acetylated with acetic anhydride ( l ml) and anhydrous pyridine ( l . m l ) . The product was i s o l a t e d i n the usual way and c r y s t a l l i s e d from ether-petroleum ether (b.p. 30-60°) , m.p. 82°l M S * * -43° (c, 1.3 i n chloroform). - 164 - Experimental Section B 3 , 4 ,6-Tri - 0-acetyl-D-galactal (40) (a) D-galactose (55 g) was added portionwise over 45 minutes, with control of temperature between 3 5 - 4 0 ° , to a s t i r r e d solution of acetic anhydride (200 ml) containing 70$ perchloric acid (1 .2 ml), and the solution was stood overnight at room temperature, or at 40° for 2 hours. Red phosphorus (15 g) was added to the s t i r r e d .solution, followed by the dropwlse addition of bromine (29 ml) and then of water (15 ml), the temperature being kept through- out at or below 20° by cooling i n an ice-water bath. The mixture was then warmed to room temperature and allowed to stand for 3 hours, then f i l t e r e d with suction, the re- sidue being washed with a l i t t l e g l a c i a l acetic acid. The amber coloured solution containing 2 , 3 , 4 , 6 - t e t r a - 0 - a c e t y l - oc-D-galactopyranosyl bromide was Immediately added drop- wise to a zinc-acetic acid reduction mixture containing sodium acetate and copper sulphate, maintained at -5 to -10 prepared as described under 3 , 4-di-O-acetyl-D-xylal; sub- sequent reaction and i s o l a t i o n of crude 3 , 4 , 6 - t r i - O - a c e t y l - D-galaetal was also ca r r i e d out as described f o r the i s o l a - t i o n of the pental. The crude product was p u r i f i e d by frac t i o n a l d i s t i l l a t i o n under high vacuum. The f r a c t i o n b.p. 135-139°/0.5 mm (46 g) c r y s t a l l i s e d slowly on standing i n - 165 - the r e f r i g e r a t o r ; [cx]^2 - 1 5 ° (e, 7 .0 i n chloroform) 3 n^ 0 1.4671; one zone by T.L.C. (benzene-methanol, 96:4 v/v). (b) 2 , 3 * 4 , 6-Tetra - 0-acetyl-cx-D-galactopyranosyl bromide (89) was i s o l a t e d during the preparation of 3 , 4 , 6 - t r i - O - acetyl-D-galactal on one occasion. Chloroform (150 ml) was added to the hydrobromination reaction mixture a f t e r standing at room temperature for 2 hours, and the mixture was f i l t e r e d through a layer of glass wool, the reaction f l a s k and f i l t e r funnel being washed with an additional 30 ml of chloroform, and vigorously extracted with two portions (400 ml and 150 ml) of i c e - c o l d water. The chloro- form layer was added to a s t i r r e d saturated aqueous solution of sodium hydrogen carbonate i n a 1 I beaker, and the mix- ture was transferred to 'a separatory funnel and thoroughly shaken. The separated chloroform layer was s t i r r e d . f o r 10 minutes with dry s i l i c i c acid (5 g), f i l t e r e d and evaporated under reduced pressure to a pale yellow syrup. This was dissolved i n ether (300 ml) and shaken with charcoal (5 g), calcium chloride ( 5 g) and sodium hydrogen carbonate (0.5 g), f i l t e r e d , and the solvent removed under reduced pressure u n t i l about 75 ml of ether remained. On standing i n the r e f r i g e r a t o r the product c r y s t a l l i z e d , and was r e c r y s t a l l i s ed from ether-petroleum ether (b.p. 30-60°). Y i e l d 89 g of ( 8 9 ) , m.p. 82-83°. - 166 _ In cases where attempted d i s t i l l a t i o n of crude 3 * 4 , 6 - tri-O-acetyl-D-galactal under reduced pressure resulted i n decomposition, p u r i f i c a t i o n was effected by chromatography on a column of P l o r i s i l , prewashed with anhydrous benzene, using benzene-methanol (10G'2, v/v) as developing solvent. The f i r s t zone to be eluted (one spot by T.L.C.) was i d e n t i f i e d as 3 * 4 , 6-tri-0-acetyl-D-ga.lactal by comparison of i t s infrared and n.m.r. spectra with those of an authentic sample. Reaction of 3,4 , 6-Tri-O-acetyl-D-galactal with Carbon Monoxide and Hydrogen Reaction conditions and product i s o l a t i o n procedures were i n general si m i l a r to those employed previously with 3,4-di-O-acetyl-D-xylal. A solution of 3 , 4 , 6 - t r i - O - a c e t y l - D-galactal (15 g) and dicobalt octacarbonyl (3 .5 g) i n an- hydrous benzene (50 ml) was shaken with carbon monoxide (1300 psi) and hydrogen (1900 psi) i n a high pressure re- action vessel at 13O-I35 0 f o r 2 I/2 hours. On cooling to room temperature the decrease i n pressure was approximately equivalent to the absorption of 3 moles of gas (C0+2Hg). After pressure was released, the reaction mixture was trans- ferred to a column (14 x 8 cm diam.) of F l o r i s i l and cata- l y s t was eluted with petroleum ether (b.p. 30-60°). E l u t i o n with benzene-ethanol (9 :1* v/v) and evaporation of solvent gave a syrup ,(.13.5 g) • - 167 - Beacetylation with 0.1N sodium methoxide i n methanol at room temperature f o r 18 hours, n e u t r a l i s a t i o n with s o l i d carbon dioxide and evaporation of solvent gave a water- soluble product which, a f t e r deionisation with Amberlite IR-120 (H +) r e s i n and freeze-drying, afforded a p a r t i a l l y c r y s t a l l i n e product consisting p r i n c i p a l l y of a mixture of two polyols i n approximately equal amounts. I s o l a t i o n of the two main components of the mixture was c a r r i e d out by preparative paper chromatography to give Fraction A (Rp 0.24) and Fraction B (R„ 0 . 2 1 ) . From an amount of 0.43 g of crude mixture, 0.16 g of A and 0.14 g of B were i s o l a t e d . Characterisation of Fractions A and B Fraction A (2 ,6-anhydro -3-deoxy-B-galacto-heptitol (91)) Fraction A, Rp 0.24, was dissolved i n methanol, c l a r i f i e d and decolourised by f i l t r a t i o n through a layer of Celite-Darco 60 , and c r y s t a l l i s e d by the addition of isopropyl ether to t u r b i d i t y ; m.p. 1 5 8 - 1 5 9 ° ' M p 7 + 2 4 ° (c, 0 . 8 i n water). Calc. f o r C^H .^0^ c , 4 7 . 1 8 , H, 7 . 9 2 . Found1* C, 4 7 . 5 0 ; H, 8 .07. , N.m.r. signals (BgO) : multiplet 3.25 - 4'. 13 ( 8 ) ; multiplet 1.32 - 1 .87 ppm ( 2 ) . - 168 - 2 , 6-Anhydro - 3-deoxy-l,4 , 5 , 7-tetra-O-p-nitrobenzoyl- D-galacto-heptitol Fraction A (35 mg) was heated with p-nitrobenzoyl chloride (0 .30 g) i n anhydrous pyridine (1.5 tnl) at 90° f o r 1 hour. I s o l a t i o n of the..product i n the usual way gave a s c l i d (95 mg) which was r e c r y s t a l l l s e d from ethyl acetate- petroleum ether (b.p. 30-60°); m.p. 2 1 0 - 2 1 1 ° : M j j 3 - 1 2 ° (c, 2 .0 i n chloroform). Calc. f o r C^H^O-j^H^: C, 5 4 . 2 8 ; H, 3 . 3 8 ; N, 7 . 2 3 . Found: C, 54.6.1; H, 3-5©; N, 7.47- Fraction B ( 2 J 6-anhydro - 3-deoxy-D-talo-heptitol (90)) The slower-moving component, R p 0 . 2 1 , was c r y s t a l l i s e d i n a s i m i l a r way from methanol-isopropyl ether; m.p. l68°# LP3D +68° (c, 1.1 i n water). Calc. f o r C^H^O^: C, 4 7 . 1 8 ; H, 7 .92. Found: C, 4 6 . 9 4 ; H , 8 .22. N.m.r. signals (D 2 0 ): multiplet 3.36 - 4 .33 ( 8 ) ; multiplet 1.53 - 2.13 ppm ( 2 ) . 2 , 6-Anhydro - 3-deoxy-l,4 , 5 , 7-tetra - 0-p-nitrobenzoyl- D-talo-heptitol A portion of the c r y s t a l l i n e Fraction B was converted to a tetra - 0-p-nitrobenzoyl derivative i n the usual way to give a s o l i d which was r e c r y s t a l l i s e d from chloroform- petroleum ether (b.p. 3 O - 6 0 0 ); m.p. 1 2 9 - 1 3 0 °(softening at - 1 6 9 - 1 1 3 - 1 1 5 ° ) J [pc] + 2 3 (c, 0 . 8 i n chloroform). Cale. for C 3 5 H 2 6 ° 1 7 N 4 : G > 5 4 . 2 8 ; H, 3-38; N, 7 . 2 3 . Found: C, 5 4 . 5 9 5 H, 3 . 4 5 J N, 7 . 7 0 . 2,6-Anhydro -3-deoxy -4 ,5-O-isopropylidene-D-talo- - 1 1 1 1 1 1 1 1 1 1 , I I h e p t i t o l ( 1 0 0 ) Fraction B ( 5 5 wig) i n anhydrous acetone (2 ml) con- tai n i n g 4 $ sulphuric acid was s t i r r e d at room temperature for 20 hours with exclusion of moisture. Neutralisation with 5N sodium hydroxide solution, f i l t r a t i o n to remove sodium sulphate and evaporation gave an o i l . This was ex- haustively extracted with b o i l i n g carbon te t r a c h l o r i d e , which on evaporation under reduced pressure gave a c r y s t a l - l i n e residue of the mono-isopropylidene derivative (100) (23 mg). Further treatment of the CClf|- insoluble residue with a c i d i f i e d acetone as described above afforded an ad d i t i o n a l amount (21 mg) of the der i v a t i v e . The combined products were r e c r y s t a l l i s o d from carbon te t r a c h l o r i d e ; m.p. 1 0 4 - 1 0 5 ° ; [ocjg 1 +12° (c, .1.6 In chloroform). Calc. fo r C 1 0 H l 8 0 : C, 5 5 . 0 3 ; H, 8 . 3 1 . Found: C, 5 4 . 8 2 ; H, 7 . 9 7 . - 170 - 2 1 6-Anhydro - 3-deoxy-4 ,5-O-isopropylidene - 1 , 7-di - 0 - (p-tolylsulphonyl)-D-talo-heptltol (106) To a solution of the mono-lsopropylidene derivative (100) (33 mg) i n dry pyridine (1 .0 ml) was added p-toluene- sulphonyl chloride (66 mg). After standing for 18 hours at room temperature i n a stoppered f l a s k the solution was s t i r r e d with water (0 .1 ml) fo r 20 minutes. More water was added u n t i l the solution became turbid, and on standing c r y s t a l s (50 mg) of the dl-O-p-tolylsulphonyl derivative formed. The product was r e c r y s t a l l i s o d from methanol; m.p. 1 3 5 ° ; M p 3 -42° (c, 1.7 i n chloroform). Calc. for C 24H " 3 0 0 9S 2: C, 5^.78; H, 5-74. Found: C, 5 5 . 0 8 ; H, 6 . 0 1 . Reaction of (106) with Sodium Iodide (a) When a solution of the di-O-p-tolylsulphonyl d e r i - vative (106) (8.4 mg) and sodium iodide (25 mg) i n acetone (0.4 ml) was heated i n a sealed tube at 118° f o r 26 hours, sodium p-toluenesulphonate was preci p i t a t e d i n an amount (6.4 mg) equivalent to the replacement of both tosyloxy groups of (106) by iodide. (b) When the sealed tube reaction ( 8 . 0 mg of (106) i n acetone (0 .5 ml) containing sodium iodide (23 mg)) was car r i e d out at 100°, an appreciable quantity of sodium p- toluenesulphonate was preci p i t a t e d within minutes. After 25 minutes the amount i s o l a t e d (2 .8 mg) was approximately - 171 - equivalent to the replacement of only one tosyloxy group of ( 106) . Consumption of Periodate Ion Fraction A Absorbance readings at 223 tflyw were measured at i n - tervals (Beckmann DU, 1 cm c e l l s ) of an aqueous solution containing 0.457 x 10_2tM of Fraction A and 0.917 x 1 0"V of sodium periodate (Reading A). Simultaneous readings 4 were taken of a solution containing 0.457 x 10 M of Frac- t i o n A (B)i and of a solution containing 0.917 x 10~^M of sodium periodate (C). Time (B +C)-A( a) (B+C)-A( b) C o Moles periodate/ •mole substrate 1 min 0 .289 0.312 O.63 47 min 0.448 0 .483 0 .97 17 hours 0.401 0 .433 0 .87 (a) : decrease i n absorbance due to consumption of periodate (b) : f r a c t i o n of known amount of periodate consumed Fraction B By a s i m i l a r procedure, but at a 1 5-fold d i l u t i o n , corresponding values obtained with Fraction B were - 172 - Time (hrs) (B+C)-A (B+C)-A C o Moles periodate/ mole substrate 1 0.463 0.473 0.88 2 0.477 . 0 . 4 8 7 0.91 9 0.478 0.490 0.92 23 0.503 O.51.I 0.95 Enantiomeric 2-Deoxy - 3 - 0-(l , 3-dlhydroxy - 2-propyl)- g l y c e r o - t e t r i t o l s (94) and (95) Fraction A (53 mg) was dissolved i n a solution (5 ml) containing periodic acid (80 mg, 5©$ excess), and the course of the oxidation was followed by observing the change i n o p t i c a l rotation of the solution, contained i n a 2 dm polarimeter tube protected from l i g h t . After 2 hours the solution was neutralised with barium carbonate and f i l t e r e d into a solution of sodium borohydride (5© mg) i n water (3 ml). After 90 minutes at room temperature the solution was neutralised with acetic acid, deionised by s t i r r i n g with Amberlite IR-120 (H +) r e s i n , f i l t e r e d and evaporated to a s o l i d , which was repeatedly evaporated with r -125 methanol to afford a syrup (52 mg); L ° y D + 2 1 ° (c, 3 .7 i n water). N.m.r. signals (D2o): multiplet 3 .50 - 3 . 8 5 , with sharp signal at 3.68 ( l 0 ) j multiplet (apparent quartet) 1.47 - 1.92 ppm ( 2 ) . - 1 7 3 - The dextrorotatory t e t r o l ether was characterised as the tetra-O-p-nitrobenzoyl derivative; a portion ( 2 1 mg) was heated with p-nitrobenzoyl chloride ( 0 . 2 g) i n pyridine ( l ml) and the product was i s o l a t e d i n the usual manner as a syrup ( 8 5 mg) which s o l i d i f i e d slowly on stand- ing under methanol at room temperature, and was r e c r y s t a l l i s e d n r -i24 o , , . from ethyl acetate; m.p. 1 5 0 - 1 5 1 0 ; L P < ] D + 2 3 (c, 0 . 9 i n chloroform). Calc. f o r C 3 5 H 2 8 O 1 7 N 4 ! c , 5 4 . 1 3 ; H, 3 . 6 3 ; N, 7 . 2 1 . Found: C, 5 4 . 4 6 ; H, 3 - 5 5 J N, 7 . 2 9 . Fraction B ( 4 9 mg) was s i m i l a r l y converted by perio- date oxidation followed by sodium borohydride reduction, with i s o l a t i o n of the product as described above, to a levorotatory t e t r o l ether ( 4 9 mg); [p<]|p - 2 3 ° (c, 3 . 2 i n water), whose n.m.r. spectrum.(D^O) was i d e n t i c a l with that described above for the dextrorotatory enantiomer. A portion ( 2 1 mg) of the product was converted i n the usual manner to a tetra-O-p-nitrobenzoyl derivative which was c r y s t a l l i z e d as described above; m.p. 1 5 0 - 1 5 1 0 ; - 2 3 ° (c, 1 . 1 i n chloroform). Calc. for C^HggO-jjNjj *• ' C, 5 4 . 1 3 J H, 3 . 6 3 . Founds C, 5 4 . 5 5 ; H, 3 . 7 1 . The inf r a r e d spectra of the two tetra-p-nitrobenzoates were i d e n t i c a l . - 174 - Attempted Syntheses of O p t i c a l l y Pure T e t r o l Ethers (a) Attempted Synthesis of L-Isomer (95) from 2,5-Anhydro- — D-glucose (117) : - '"<"•' • • r o ' D-Arabino-tei;raacetpxy-l-nltro-l-hexene (12$ To a suspension of D-arabinose (25 g) i n anhydrous methanol (50 ml) and anhydrous nitromethane (90 ml) i n a 500 ml, 3-necked f l a s k f i t t e d with an e f f i c i e n t mechanical s t i r r e r was added a solution of sodium (5*25 g) In anhydrous methanol (175 ml). The mixture was s t i r r e d at room tempera- ture for 20 hours .with exclusion of moisture, and the re- s u l t i n g s o l i d mass of sodium ac l - n l t r o a l c o h o l s was collected by f i l t r a t i o n and washed with a small volume of cold metha- nol and with petroleum ether (b.p. 30-60°). The s o l i d re- sidue was immediately dissolved i n i c e - c o l d water (200 ml) and deionised by passage through a column containing Dowex- 50 t"H+) r e s i n (200 ml). The effluent and washings ( t o t a l 500 ml) were concentrated under reduced pressure to a brown syrup, which was twice evaporated to dryness with absolute ethanol. On standing over PgO^ t n e residue S o l i d i f i e d to a c r y s t a l l i n e mass which was washed with cold ethanol to give a mixture of nitroalcohols (119) (16.5 g). I.R. ( l i q u i d f i l m ) : 1550 cm"1 (s), 1365 cm"1 (m) (NOg st r e t c h i n g ) . - 175 - The mixture of nitroalcohols (l6.5 g) was dissolved i n acetic anhydride (200 ml) containing 2 drops of concentrated sulphuric acid, and the solution was heated f o r 1 hour on a steam bath. I s o l a t i o n of the product i n the usual way gave a syrupy mixture of acetylated nitroalcohols (120). I.R. ( l i q u i d f i l m ) : 1565 cm"1 (s), 1375 cm"1 (s) (N0 2 stretching). The product was dissolved i n benzene (600 ml), sodium hydrogen carbonate (50 g) was added and the mixture was refluxed for 2 hours. F i l t r a t i o n of the cooled mixture and removal of solvent under reduced pressure gave a pale yellow c r y s t a l l i n e mass. Re c r y s t a l l J s a t i o n from ethanol gave D-arabino-tetra- acetoxy-l-nitro-l-hexene (121), (16.5 g)J m.p. 113-115°. I.R. (Nujol): 1510 cm"1 (m), 1350 cm"1 (m) (NOg i n con- jugation with C=C). 2-Acetamido-l ,2-dideoxy-l-nitro-D-mannitol (122) To D-arablno-tetraacetoxy-l-nltro-l-hexene (15 g) i n a 500 ml f i l t e r f l a s k was added methanol (150 ml); the mixture was cooled i n ice and saturated with anhydrous ammonia, when the n i t r o o l e f i n dissolved. After warming to room temperature over 8 hours with protection from moi- sture, the solvent was evaporated i n a stream of dry n i t r o - gen. The residue was f i l t e r e d with the aid of cold absolute ethanol and r e c r y s t a l l l s e d from ethanol to afford 2- acetamido-l , 2-dideoxy-l-nitro-D-mannitol (122) (4.4 g); - 176 - m.p. 1 7 1 - 1 7 2 ° . A mixture of (122) and Its epimer (123) was obtained from the mother l i q u o r s . 2-Amino-2-deoxy-D-mannose hydrochloride ( l l 6 ) A solution of (122) (4.4 g) i n 2N sodium hydroxide solution (10 ml) was added dropwlse at room temperature to concentrated hydrochloric acid (9 ml) with vigorous s t i r r i n g . The r e s u l t i n g solution was brought b r i e f l y to b o i l i n g point, cooled to 0 ° , saturated with hydrogen chloride, and f i l t e r e d to remove sodium chloride. After d i l u t i o n with water (10 ml) the solution was f i l t e r e d through a layer of C e l i t e - Darco 60 and concentrated under reduced pressure to a syrup, which was stood overnight under vacuum, over KOH, to remove re s i d u a l hydrogen chloride. The product was crystallisfed by dissol v i n g i n methanol (10 ml) containing 2-3 drops of water and adding acetone to t u r b i d i t y . Scratching, cooling and periodic additions of acetone afforded 2-amino-2-deoxy-D-mannose hydrochloride (3 .2 g) '> L —1 pp n "3-5 (c, 4 . 3 i n water). Attempted Preparation of 2,5-Anhydrd-D-glucose (117) To a solution of 2-amino-2-deoxy-D-mannose hydrochloride (3 g) i n water (50 ml) was added mercuric oxide (16 g) and the mixture was heated on a b o i l i n g water bath for 3© minutes. The cooled mixture was f i l t e r e d , saturated with hydrogen - 177' - sulphide, r e f i l t e r e d through Gelite-charcoal and evaporated under reduced pressure to a syrup. This did not c r y s t a l l i s e on d i s s o l v i n g i n a small volume of methanol; considerable darkening of the product ensued on standing and only a brown amorphous s o l i d was obtained. (b) Attempted Synthesis of D-isomer (94) from 2,5-Anhydro- D-mannose (115) . 2,5-Anhydro-D-mannose (115) A solution of 2-amino-2-deoxy-D-glucose hydrochloride (15 g) i n water (100 ml) was cooled i n an i c e - s a l t bath u n t i l a quantity of i c e had formed i n the sol u t i o n . To t h i s was added sodium n i t r i t e ( 6 . 0 g) dissolved .in i c e - c o l d water (35 ml) containing g l a c i a l acetic acid ( l ml). xThe r e s u l t i n g solution was,kept near i t s freezing point for 2 ^ / 2 hours and then stood i n the r e f r i g e r a t o r f o r 24 hours. G l a c i a l acetic acid (1.5 ml) was added and the solution was vigorously aerated at room temperature f o r 30 minutes to remove nitrous acid, and then evaporated to o a mobile syrup under reduced pressure at 25 . The product was dehydrated by several extractions with 10-15 ml portions of anhydrous acetone, dissolved i n methanol, f i l t e r e d and evaporated to a syrup, which was twice evaporated to dryness under reduced pressure at 25° with anhydrous benzene. The r e s u l t i n g 2,5-anhydro-D-mannose (10.5 g) was characterized as'ithe p-nitrophenylhydrazone, m.p. 1 7 9 - 1 8 1 ° . - 178 ,- Attempted Preparation of Acetylated N i t r o o l e f i n To a s t i r r e d solution of 2,5-anhydro-B-mannose (10 g i n anhydrous methanol (30 ml) and anhydrous nitromethane (4Jp ml) i n a 250 ml, 3-necked f l a s k was added a solution of sodium (2.5 g) i n anhydrous methanol (70 ml), with immediate formation of a white s o l i d . After s t i r r i n g at room temperature f o r 3 hours with exclusion of moisture the s o l i d was co l l e c t e d by f i l t r a t i o n and lushed with a small volume of cold methanol and with^fpetroleum ether (b.p. 30-60°). The s o l i d was dissolved i n ice- c o l d water (100 ml) and deionised by passage through a column.of Dowex-50 (H +) r e s i n . Evaporation of the combined effluent and washings under reduced pressure gave a syrup which was dried by azeotroping with benzene-ethanol. The product (7.5 g) did not c r y s t a l l i s e on standing over P2°5* I , R > ( l i q u i d f i l m ) : 1550 cm"1 (s), 1365 cm"1 (m) (N0 2 s t r e t c h - ing)* T.L.C. (water-saturated ethyl methyl ketone); two closely-moving zones. The mixture of nitro-alcohols (7.5 g) was dissolved i n acetic anhydride (100 ml) containing 2 drops of sulphur! acid and the solution was heated on a steam bath f o r 1 V2 hours and stood overnight at room temperature. I s o l a t i o n of the product i n the usual way gave a syrup (10.0 g); I.R. ( l i q u i d f i l m ) : 1565 cm"1 (s), 1375 cm"1 (s) (N0g. stretc h i n g ) . This was dissolved i n benzene (200 ml) and - 179}- the solution was refluxed with sodium hydrogen carbonate (20 g) f o r 11 hours. The Infrared spectrum of the syrupy product Isolated by f i l t r a t i o n and evaporation of solvent was unchanged from that of the mixture of acetylated n i t r o - alcohols. No change i n the in f r a r e d spectrum was observed on further r e f l u x i n g i n benzene solution with anhydrous sodium acetate. 2-Deoxy-3-0-(l ,3-'dihydroxy-2-propyl) - L - g l y c e r o - t e t r i t o l (95) from 2 ,6-Anhydro-3-deoxy-D-gluco-heptltol ( l 3 0 ) ( a ) Oxidation of 2 ,6-anhydro-3-deoxy-D-gluco-heptitol (130) with an excess of periodic acid, followed by reduction of the r e s u l t i n g dialdehyde with an aqueous solution of sodium borohydride i n water, and .isolation of the product as described previously gave 2-deoxy - 3 - 0-(l , 3-dihydroxy - 2 - propyl) - L - g l y c e r o - t e t r i t o l (95); [oc] 2 ) 2 + 2 6 0 (c, 2 .9 i n water). The t e t r o l ether (95) formed a t e t r a - 0 - p - n i t r o - benzoyl derivative, m.p. 1 5 1 - 1 5 2 ° ; [P<D§ 1 +22 0 (c, 1.2 i n chloroform). The melting point of t h i s derivative was undepressed on admixture with the corresponding derivative of the dextrorotatory t e t r o l ether prepared from Fraction A, ( + 2 3 ° ) , and the in f r a r e d spectra of the two p- nitrobenzoates were i d e n t i c a l . (a) The structure of 2 ,6-anhydro -3-deoxy-D-glueo-heptltol ( l30) had been established by c o r r e l a t i o n w i t h - 4 , 5 , Y - t r i - 0 - a c e t y l - 2 , 6 - anhydro-3-deoxy-l-0-p-bromobenzenesulphonyl-D-gluco7heptitol fi (129 ) 1 5 5 1 whose -structure had been proved by X-ray a n a l y s i s 1 - ^ t - 180 - Experimental Section C Hydroformylatlon of 3,4-di-O-acetyl-D-xylal 3,4-Di-O-aeetyl-D-xylal (12.0 g) and dicobalt octa- carbonyl (3 g) were dissolved i n anhydrous benzene; the volume of the solution (60 ml), contained i n the glass l i n e r of a high pressure reaction vessel, gave a net void of 200 ml. After flushing with carbon monoxide, additional carbon monoxide was added to a pressure of 600 p s i , followed by hydrogen (2400 ps i ) J on e q u i l i b r a t i o n the i n i t i a l gas pressure was 2930 p s i at room temperature. The bomb was then heated with shaking to a temperature of 1 1 5 ° . The pressure increased to a maximum of 3730 p s i a f t e r 35 min- utes from the attainment of a constant temperature, and then began to f a l l . After an addit i o n a l 30 minutes the reaction vessel and contents were ra p i d l y cooled to room temperature by immersion i n i c e , when the pressure had de- creased by 220 p s i , to 2710 p s i , equivalent to the absorp- t i o n of 2 moles of synthesis gas (H g+ GO). Unreacted gas pressure was released, the reaction mixture was transferred to a column of F l o r i s i l and catalyst was eluted with pet- roleum ether (b.p. 30-60°). E l u t i o n with benzene-isopropyl alcohol (9 J1* v/v) and removal of solvent then gave a syrup (10.0 g). An add i t i o n a l quantity of a mobile syrup (2 .5 g) was subsequently recovered by f i l t r a t i o n and - 181 - evaporation: of the petroleum ether f r a c t i o n , a f t e r decomposi- t i o n of catalyst on standing at room temperature. The l a t t e r f r a c t i o n , i d e n t i f i e d as 3 , 4-di-O-acetyl-D-xylal (I.R. spec- trum and T.L.G. (benzene-methanol, 96*4, v/v) alongside an authentic specimen) c r y s t a l l i s e d on standing i n the re- f r i g e r a t o r . T.L.C. of the main f r a c t i o n (benzene-methanol, 9 5 : 5 . v/v) alongside 3 , 4-di-O-acetyl-D-xylal and the mixture of di-O-acetyl-anhydrodeoxy-hexitols previously obtained by the hydroxymethylation of ( 2 2 ) , showed two well separated zones corresponding to both controls. Chromatography of a portion (1.5 g) of the main f r a c t i o n on a column of neutral alumina, using benzene-methanol ( 9 5 : 5 , v/v) as developing solvent, gave 3 , 4-di-O-acetyl-D-xylal (0.55 g) and a mixture of reaction products (0 .93 g). The l a t t e r f r a c t i o n showed a n.m.r. signal (CCl^) at & = 9«35 ppm (CHO) (intensity r e l a t i v e to acetate absorption around 8 = 2 ppm consistent with approximately 15$ di-O-acetyl- deoxyanhydro-aldehydo-hexoses (133) and ( 1 3 4 ) ) . Reaction with 2 ,4-Dinitropheny.lhydrazine A portion (1.2 g) of the reaction mixture eluted from F l o r i s i l with benzene-iso.propyl alcohol was dissolved i n ethanol (20 ml) containing 2-3 drops of acetic acid, and the solution was heated to b o i l i n g point on a steam - 1821- bath. To t h i s was added portlonwise a hot, saturated solution of 2,4-dinitrophenylhydrazine'in. ethanol, u n t i l the orange colour Imparted to the solution no longer faded to yellow. On d i l u t i n g with water to t u r b i d i t y , and standing i n the r e f r i g e r a t o r , a yellow s o l i d (0.39 g) separated, which was col l e c t e d by f i l t r a t i o n and dried over calcium chloride. T.L.C. (benzene-methanol, 95:5* v/v); 2 c l o s e l y moving components, Fraction X (faster-moving) and Fraction Y (slower-moving), plus traces of others; no additional zones revealed with s u l p h u r i c - n i t r i c acid spray reagent. N.m.r. signals (CDCl^): singlet 11.07 ( l ) , N-NH-; doublet (J = 3 c/s) 9.04 ( l ) , aromatic H^J quartet 8.30 ( l ) , aromatic RV; pair of overlapping doublets (J = 9 C$te), 7.87 and 7.95 ( l ) , aromatic Hg; doublet (J = 6 cjfs) 7.58 ( 1 ) , N CH-; multiplet 3 .15-5.35 (5)5 3 sharp signals, at 2 . 0 7 , 2.12 and 2 . l 6 ppm ( 6 ) , OCOCH^, superimposed on multiplet ( 2 ) , C-CH2-C. Fraction Y ( 4 , 5-Dl - 0-acetyl - 2 , 6-anhydro - 3-deoxy- aldehydo-D-lyxo-hexose 2 ,4-dinitrophenylhydrazone) When the mixture of derivatives from the reaction with 2 , 4-dinitrophenylhydrazine was t r i t u r a t e d with warm ethanol (2-3 ml), p a r t i a l d i s s o l u t i o n occurred and a granular yellow s o l i d separated. This was iso l a t e d by f i l t r a t i o n , washed with a l i t t l e cold ethanol, and - 183 - r e c r y s t a l l i z e d as fine yellow needles from chloroform- hexane; m.p. 225-226°; L°il|2 -6©° (e, 2.5 In chloroform). One spot by T.L.C. (benzene-methanol, 95l5, v/v) corre- sponding to slower-moving of two components present i n mixture. N.m.r. ( C D C l ^ ) d i f f e r e d from mixture of X and Y described above i n showing doublet (J = 9 4/&) at 7.85 ( l ) , aromatic C-6 proton, and two sharp signals 2.13 and 2.17 ppm (6), OCOCH3. Gale, for C 1gH 1gN ]jO g: C, 4,6.83> H, 4.42; N, 13.66. Pound: C, 46.60; H, 4.64; N, 13-77. 1,5-Anhydro-4-deoxy-D-arabino-hexitol (44) from Fraction Y An additional amount (0.9 g) of Fraction Y was pre- pared by reaction of the main f r a c t i o n from the hydroformy- l a t i o n of (22) (6.8 g) with 2,4-dinitrophenylhydrazine, and i s o l a t i o n as described above. A portion (0.4 g) i n chloro- form (3 ml) was added to a mixture consisting of fre s h l y d i s t i l l e d pyruvic acid (4 ml) and a 10$ solution of hydrogen bromide i n acetic acid (0.2 ml). The solution was heated o / at 40-50 for 1 hour and then stood overnight at frDom temp- erature. A yellow c r y s t a l l i n e s o l i d (0.2 g) was is o l a t e d by f i l t r a t i o n , washed with a small volume of cold chloroform, and i d e n t i f i e d as pyruvic acid 2,4-dinitrophenylhydrazone (m.p., I.R. spectrum). The f i l t r a t e was extracted with two 10 ml portions of chloroform, and the extract was washed with three 10 ml portions of saturated sodium hydrogen car- bonate solution to remove dissolved pyruvic acid 2,4-dinitro- - 184 - phenylhydrazone, and then with water. After drying over magnesium sulphate and f i l t e r i n g , solvent was removed under reduced pressure to afford a syrup (110 ragf. recovery not quan t i t a t i v e ) . N.m.r. (CCl^) : singlet at 9.35 ppm (CHO). I.R. ( l i q u i d f i l m ) : carbonyl bands at 1740 cm"1 (ester) and 1700 cm"1 (aldehyde); band at 3400 cm"1 (OH). When a portion (20 mg) of the aldehydo-compound was reacted i n ethanol solution with 2,4-dinitrophenylhydrazine as described previously, a c r y s t a l l i n e product separated on cooling which was i d e n t i f i e d as Fraction Y (m.p., T.L.C, n.m.r.). A portion (35 mg) of the aldehydo-product from the exchange reaction with pyruvic acid was dissolved i n metha- nol (2 ml) and added dropwise to a solution of sodium boro- hydride (30 mg) i n water (2 ml). After standing overnight at room temperature the solution was neutralised by the addition of acetic acid and deionised by passage through a small column of Amberllte IR-120 (H +) r e s i n . The com- bined effluent and washings were evaporated to a s o l i d residue which was repeatedly -evaporated to dryness with methanol to afford a syrup (8 mg). One spot on paper chromatography, alore and when superimposed on 1,5-anhydro- 4-deoxy-D-arabino-hexitol (44), two spots when superimposed on the slower-moving L-xylo-isomer (45). - 185 - Fraction X (4,5-Pi-0-acetyl-2,6-anhydro-3-deoxy-aldehydo- D-xylo-hexose 2 ,4-dinitrophenylhydrazone) The alcohol-soluble portion of the mixture of hydra zones remaining a f t e r i s o l a t i o n of Fraction Y as .described above was evaporated to a syrup. A portion (45 mg) was applied i n chloroform solution as a narrow streak near the lower edge of a plate (20 cm wide x 45 cm long x 0 .8 mm thick) of s i l i c a gel G, and fractionated by multiple ascending development, using chloroform as developing solvent, with a i r drying be- tween successive developments. The faster-moving of the two major components was cut out, eluted with chloroform- ethanol ( l : l , v/v), f i l t e r e d and evaporated to a syrup t which was c r y s t a l l i s e d from chloroform-hexane as f i n e needlesj m.p. 1 3 2 ° ; M D - 16° (c, 0.4 In chloroform). N.m.r. (CDC13) : one aeetoxy signal at 2.05 ppm. Gale, for C^gH^gN^O^: C, 46.835 H, 4.42. Found: C, 46.901 H, 4 . 1 0 . Dl -0 -a c e t y l -a nhydrodeoxy -a'ldehydo -hexos es (133) and (134) by Oxidation of Di-O-acetyl-anhydrodeoxyhexitols (74) and (75) (a) To a portion (0 .88 g) of the syrupy product from the hydroxymethylation of 3 j 4-di - 0-acetyl-D-xylal (section A), In anhydrous pyridine (5 ml) was added p-toluenesulphonyl chloride (1 .0 g), and the mixture was stood at room temperature - 1&6 - for 20 hours. I s o l a t i o n of the product i n the usual way gave a syrup (1 .0 g) containing the 6 - 0-p-tolysulphonyl derivatives (76) and (77) . I.R. ( l i q u i d f i l m ) ; hands at —1 1 l l 8 0 cm" (S=0 stretching) and 1600 cm (aromatic); no hydroxyl absorption. The product (l.O g) i n anhydrous d i - methylsulphoxide (5 ml), was added dropwise to a mixture of dimethylsulphoxide (20 ml) and sodium hydrogen carbonate (3 g)> through which was passing a stream of nitrogen,.and which was maintained at 150° i n an o i l bath. After 5 minutes the mixture was cooled, f i l t e r e d , and solvent was removed under reduced pressure to afford a brown syrup, whose infrared spectrum was unchanged from that described above. (b) To a solution of the hydroxymethylation product comprising di - 0-acetyl-anhydrodeoxyhexitols (7*0 and (75) (0.37 g) i n anhydrous dimethylsulphoxide (6 ml), was added anhydrous phosphoric acid (0 .1 ml) and N,N'-dicyclohexyl- carbodiimide (1 .8 g). After standing at room temperature under anhydrous conditions for 24 hours the mixture was f i l t e r e d , and the c r y s t a l l i n e residue of N,N'-dicyelohexyl- urea was washed with absolute ethanol to give a t o t a l volume ( f i l t r a t e +washings) of 30 ml. To a portion (15 ml) of t h i s s o l ution, heated on a steam bath, was added a hot, saturated solution of 2,4-dinitrophenylhydrazine i n ethanol, u n t i l an orange colour persisted on b o i l i n g . D i l u t i o n with water to - 187 - t u r b i d i t y and cooling gave a yellow s o l i d which was collected by f i l t r a t i o n and dried. Separation of 2 , 4-dinitrophenyl- hydrazones from co-precipitated N,N'-dicyclohexylurea was effected by extraction of the former with a small volume of cold chloroform, f i l t r a t i o n and evaporation to a yellow s o l i d (165 mg). Preparative scale TLC of a portion (30 mg) of the mixture of hydrazones afforded two pure components; 4,5-di-Q-acetyl-2,6-anhydro-3-deoxy-aldehydo-D-lyxo-hexose 2 , 4-dinitrophenylhydrazone (17 mg), m.p.' 225-226°, and the 2 , 4-dinitrophenylhydrazone of the corresponding D-xylo- isomer, (9 mg), m.p. 1 3 2 ° , both id e n t i f i e d , with the two 'fractions (Y and X respectively) described previously. • When the same experiment was car r i e d out with omission of the mixture of acetylated anhydrodeoxyhexitols, no pre- c i p i t a t e was Isolated following d i l u t i o n of the reaction mixture with water. S i m i l a r l y , a control experiment i n which N,N'-dicyclohexylcarbodiimide was not present did not r e s u l t i n the i s o l a t i o n of any reaction product following treatment with 2 , 4-dinitrophenylhydrazine. - 188 - REFERENCES 1. H. Akins and G. Krsek, J. Am. Chem. S o c , 70, 383 (1948). 2. I. Wender, M. Orchin and H. H. Storch, J. Am. Chem. Soc. 72, 4842 (1950). 3. F, Fischer and H. Tropsch, Brennstoff-Chem., _4, 276 (1923). 4. D. F. Smith, C. 0. Hawk and P. L. Golden, J . Am. Chem. S o c , 5 2 , 3221 (1930). 5. 0. Roelen, Ger. Patent 103,362 ( f i l e d 1938)? U.S. Patent 2 ,327,066 (1943)1 Chem. Abstracts, 38, 550 (1944). 6. I. Wender, H. W. Sternberg and M. Orchin, "Catalysis", V o l . 5 , ed. by P. H. Emmett, New York, Reinhold Publishing Corporation, 1957, p. 73 . 7. H..W. Sternberg and I. Wender, Proceedings of International Conference on Coordination Chemistry, Chem. Soc. Spec. Publn. No. 13, (1939), P. 3b. : 8. C. W. B i r d , Chem. Rev., 6 2 , 283 (1962). 9. H. Adkins and G. Krsek,* J . Am. Chem. S o c , 71 , 3051 (1949). 10. J . Halpern, Quart. Rev., 10, 463 (1956). 11. M. Orchin, L. Kirch and I. Goldfarb, J . Am. Chem. S o c , 78,-5450 (1956). — 12. A. R. Martin, Chem. and Ind., 1536 (1954). 13. G. Natta, R. E r c o l i , S. Castellano and P. H. B a r b l e r l , J . Am. Chem, S o c , 76 , .4049 (1954). 14. G. L. Karapinka and M. Orchin, J . Org. Chem., 26 , 4l87 (1961). = 15. L. Kirch and M. Orchin, J . Am. Chem, S o c , 80, 4428 (1958). = 16. I. Wender, S. Metlin, S. Ergun, H. W. Sternberg and H. Greenfield, J . Am, Chem, S o c , 78, 5401 (1956) . 17. R. F. Heck and D. S. Breslow, J , Am. Chem, S o c , 8 3 , 4023 (1961). = - 1 8 9 - 18. D. S. Breslow and R. F. Heck, Chem. and Ind., 467 ( i 9 6 0 ) . 19. T. H. C o f f i e l d , J . Kozlkowski and R. D. Closson, J . Org. Chem., 22 , 598 (1957). ' : 2 0 . L. Marko, Proc. Chem. S o c , 67 (1962). 2 1 . C. L. Aldrldge and H. B. Jonassen, Nature, 188, 4o4 ( i 9 6 0 ) . 2 2 . C. L. Aldrldge and H. B. Jonassen, J . Am. Chem. S o c , 8 5 , 886 (1963). = 2 3 . J . G. Traynham, J . Am. Chem. S o c , 78 , 4042 (1956). 24. I. Wender, J . Feldman, S. Metlln, B. H. Gwynn and M. Orchin, J . Am. Chem. S o c , 77, 5760 (1955). 2 5 . J . Falbe and F. Korte, Ber., 97, 1104 (1964). 2 6 . I. Wender, R. Levin and M. Orchin, J . Am. Chem. S o c , 72, 4375 (1950). : — = 27 . E. I. du Pent de Nemours and Co.. B r i t . Patent 6 l 4 , 0 1 0 J Chem. Abstracts, 4 3 , 4685 (1949) . 2 8 . C. W. B i r d , R. C. Cookson, J . Hudec and R. 0 . Williams, J . Chem. S o c , 410 (1963). 2 9 . P. F. Beal, M. A. Rebenstorf and J . E. Pike, J . Am. Chem. S o c , 8 1 , 1231 (1959). : 3 0 . A. L. Nussbaum, T. L. Popper, E. P. Oliveto, S. Friedman and I. Wender, J . Am. Chem. S o c , 8 l , 1228 (1959). 31 . E. Fischer and K. Zach, S i t z . ber. k g l . preuss. Akad. Wiss., 16, 311 (1913) -"CTIem. Abstracts, 8, 73 (1914). 32. R. K. Ness and H. G. Fletcher, J . Org. Chem., 28, 435 (1963). = 3 3 . E. Fischer, Ber., 47, 196 (1914). 34. E. Fischer, M. Bergmann and H. Schotte, Ber., 53 , 509 (1920). = 35 . L. D. H a l l and L. F. Johnson, Tetrahedron L e t t e r s , 2 0 , 883 (1964). — — 36. M. Karplus, J . Chem. Phys., 30, 11 (1959). - 190 - 37. B. H e l f e r i c h , Advances In Carbohydrate Chem., 7, 209 (1952). 38. 38. D. A. Prins and R. W. Jeanloz, Ann. Rev, Biochem., 17, 67 (1948). ! = 39. W. G. Overend and M. Stacey, Advances i n Carbohydrate Chem., 45 (1953). " : : 4 0 . . M. Barczai-Martos and P. Korosky, Nature, 165, 369 (1950). 41. R. E. Deriaz, W. G. Overend, M. Stacey, E. G. Teece and L. F. Wiggins, J . Chem. S o c , 1879 (1949). 42. W. G. Overend, P. Shafizadeh and M. Stacey, J . Chem, S o c , 671 (1950). ~ 4 3 . B. M. I s e l i n and T. Reichstein, Helv. Chlm, Acta, 27, 1146, 1200 (1944). ; = 44. B, He l f e r i c h , E. N. Mulcahy and H. Ziegler, Ber., 87> 233 (1954). — - 4 5 . B. He l f e r i c h and M. Gindy, Ber., 87 , 1488 (1954). 46. L. F. Pieser and M. F i s e r , Advanced Organic Chemistry, Reinhold Publishing Corp., New York, I 9 6 I , p. 310. 47 . H. S. I s b e l l and W. W. Pigman, J , Research N a t l , Bur. Standards, 22 , 397 (1939). ' 48 . W. G. Overend, P. Shafizadeh and M. Stacey, J. Chem. S o c , 992 (1951). : ~ 4 9 . F. B. Cramer, J . Franklin Inst., 253, 277 (1952). 5 0 . C. K. Ingold, Structure and Mechanism i n Organic Chemistry, Cornell University Fress, ithaca, N.Y., 1953, p. 669. 51 . J . Meinwald, Y. C. Meinwald and T. N. Baker, J . Am. Chem. S o c , 8 5 , 2513 (1963). 5 2 . W. J . Serfontein, J. H. Jordaan and J . White, Tetrahedron Letters, lB, 1069 (.1964) . 5 3 . A. Schonberg and A. Mustafa, Chem, Rev., 40, l 8 l (1947). 54. B. He l f e r i c h and E. von Gross, Ber., 8 5 , 531 (1952). 5 5 . R. C. Hockett, A. C. Sapp and S. R. Millman, J . Am. Chem. S o c , 6 3 , 2051 (1941). 56. D. H. R. Barton and R. C. Cookson, Quart. Rev., 10, 44 (1956). = 57. J . Chatt, Chem. Rev., 48_, 15 (1951). 58 . P. T. Manolopoulos, M. Mednick and N. N. L i c h t i n , J . Am. Chem. S o c , 8_4_, 2203 (1962) . 5 9 . P. A. Levene and A. L. Raymond, J . B i o l . Chem., 8 8 , 513 (1930). = 6 0 . R. U. Lemieux and B. Fraser-Reid, Can. J . Chem., 42, 532 (1964). = 61 . R. U. Lemieux and S. Levine, Can. J . Chem., 40, 1926 (1962). ' = 6 2 . A. Rosenthal. D. Read and C. Cameron, Science, 123, 1177 (1956). = 6 3 . A. Rosenthal and D. Read, Can. J . Chem., 35 , 788 (1957). 64. A. Rosenthal and D. Read, Methods i n Carbohydrate Chemistry, V o l . I I , ed. by K. L. wnistler and M. L. wolrrom, New York, Academic Press, 1963, p. 457. 6 5 . P. A. Levene and T. Mori, J. B i o l . Chem., 8 3 , 803 (1929). 66 . F. Weygand, MethodB i n Carbohydrate Chemistry, V o l . I, ed. by R. L. Wnistler and M. L. Wolfrom, New York, Academic Press, 1962, p. 182. 67. W. G. Overend, F. Shafizadeh and M Stacey, J . Chem. S o c , 1027 (1950). — ; 6 8 . A. M. Gakhokidze, Zhur. Obschei Khim., 15, 530 (1945)> Chem. Abstracts, 40, 4673 (1946). = • 6 9 . D. W. Read, B.A. Thesis, University of B r i t i s h Columbia, 1954. 7 0 . I. Wender, Ph.D. Thesis, University of Pittsburgh, 1950. 7 1 . C. T. Bishop, Methods Biochem. Anal. 10, 1 (1962). 72. G. Zemplgn, A. Gerecs and I. Hadacsy, Ber., 6 9 , 1827 (1936). = 7 3 . J . Baddiley, J . G. Buchanan, R. E. Handschumacher and. J. F. Prescott, J. Chem. S o c , 2818 (1956). - 192 - 74. J . M. Bobbit, Advances In Carbohydrate Chem., 11, .1 (1956). 75 . R. D. Guthrie, Methods In Carbohydrate Chemistry, V o l . I, ed.by R. L. Whistler and M. L. Wolfrom, New York, Academic Press, 1962, p. 435. 76. J . S. Dixon and D. Li p k i n , Anal, Chem,, 2 6 , 1092 (1954). 77 . C. E. Crouthamel, H. V. Meek, D. S. Martin and C. V. Banks, J , Am, Chem, S o c , 71 , 3031 (1949). 78, , Rules of Carbohydrate Nomenclature, J. Org. Chem., 2 8 , I 281 (1963). = 79. L. M. Jackman, Applications of Nuclear Magnetic Resonance Spectroscopy i n Organic Chemistry, Fergamon Press Ltd., London, 1 9 5 9 , p. 5 5 . 8 0 . E. von Rudloff, D. E. Stuetz and H. F. Bauer, Can, J . Chem,, J 5 , 315 (1957). 8 1 . M. L. Wolfrom and A. Thompson, Methods i n Carbohydrate Chemistry, V o l . I I , ed. by R. L. Whistler and M. L. Wolfram, New York, Academic Press, 1963, p. 6 5 . 8 2 . P. A. J . Gorln, Can, J , Chem., 38 , 64l ( i 9 6 0 ) . 8 3 . F. L. Benton and T. E. D i l l o n , J . Am, Chem, S o c , 64, 1128 (1942). : s a " 84. Our thanks are due to Dr. E. von Rudloff, National Research Council P r a i r i e Regional Laboratory, Saskatoon, Sask., for supplying a sample of. t h i s compound. 8 5 . P. A. J . Gorin, J . Org. Chem., 24, 49 (1959). 8 6 . A. S. P e r l i n , E. von Rudloff and A. P. Tulloch, Can. J. Chem., J 5 , 1504 (1957). 8 7 . A. B. Foster, W. G. Overend. M. Stacey and G. Vaughan, J. Chem, S o c , 3367 (1954). 8 8 . F. Shafizadeh, Methods In Carbohydrate Chemistry, V o l . I I , ed. by R. L. Whistler and M. Wolfrom, New York, Academic Press, 1963, p. 409. 8 9 . F. Shafizadeh and M. Stacey, J . Chem. S o c , 3608 (1952). 9 0 . C. Tamm and T. Reichstein, Helv. Chlm. Acta, 31, 1630 (1948) - 193 - 91 . A. B. Poster, W. G. Overend and M. Stacey, J.- Chem. S o c , 974 (1951). " 9 2 . W. N. Haworth, L. N. Owen and P. Smith, J . Chem. S o c , 88 (1941). 9 3 . P. A. Levene and A. L. Raymond, J . B i o l . Chem., 102, 317 (1933). = 94. J . W. H. Oldham and J . K. Rutherford, J. Am. Chem. S o c , 54, 366 (1932). ~ 9 5 . S. Peat, Advances i n Carbohydrate Chem., 2 , 38 (1946). 96 . R. Criegee, E. Buchner and W. Walther, Ber., 73 , 571 (1940). 97. C. S. Hudson, J, Am. Chem. S o c , 31 , 66 (1909) . 9 8 . B. Coxon and H. G. Fletcher, J . Am. Chem. S o c , 8 5 , 2 6 4 l (1963). = 9 9 . C. W. Smith (to S h e l l Development Co.), U.S. Patent 2 ,489,729 (1949)J Chem. Abstracts, 44 , 3036b (1950). 100. K. Ziegler, Organic Synth.,- Col. V o l . I, 314 ( l 9 4 l ) . 101. V. Migrdichian, The Chemistry of Organic Cyanogen Compounds, Reinhold Publishing uorp., New York, l 9 4 y , p. 219. 102. B.A. Nelson, E. J . Hodges and J. I. Simon, J . Org. Chem., | 1 , 798 (1956). ~ 103. R. Z e l i n s k i and K. Yorka, J . Org. Chem., 2 3 , 640 (1958). 104. L. J . Bellamy, The Infrared Spectra of Complex Molecules, Methuen and Co. Ltd., London, 1958, p. 225. 105. J . Davo 11, J . Chem. S o c , 2526 (1949). 106. R. Paul, Compt. rend., 198, 375 (1934) . 107. R. P. Z e l i n s k i , N. G. Peterson and H. R. Wallner, J . Am. Chem. S o c , 74 , 1504 (1952). 108. W. A. Bonner, Advances In Carbohydrate Chem., 6 , 251 (1951); Met hoar* i n uarbonydrate unemistry, Vu\L. II, ed. by R. L. wnistler and H. h. woirrom, New York, Academic Press, 1963, p. 465. 109. J . Houben and K. Fuhrer, Ber., 40, 4990 (1907) . - 194 - 110. C. Paal and F. Hornstein, Ber., 39 , 1361 (1906). 111. E. Fischer and K. Hess, Ber.,45, 912 (1912) . 112. C. 0. Jeremlas and C. A. Mackenzie, J . Am, Chem. S o c , 7 0 , 3920 (1948). ' : 113. C D . Hurd and W. A. Bonner, J . Am, Chem, S o c , 67 , 1972 (1945). ~ ' = 114. W. A. Bonner, J . Am, Chem, Soc,, 6 8 , 1711 (1946) . 115. C D . Hurd and W. A. Bonner, J , Am, Chem. S o c , 67, 1664 (1945). — 116. J . Yoshimura, N. Maramatsu and T. Sato, Nippon Kagaka Zasshl, 79 , 1503 (1958). 117. E. Vischer and T. Reichstein, Helv. Chim, Acta, 27, 1332 (1944). : = 118. K. Freudenberg and K. Raschlg, Ber., 6 0 , 1633 (1927). 119. R. C. Hockett and M. L. Downing, J . Am. Chem, S o c , 64, 2463 (1942). : " 120. H. Ohle and E. Dickhauser, Ber., 5 8 , 2593 (1925). 121. K. Gatzl and T. Reichstein, Helv. Chim. Acta, 21, 914 (1938). = 122. A. Rosenthal and H. J . Koch, Can, J . Chem., 42, 2025 (1964). * = 123. R. U. Lemieux, R. K. K u l l n l g , H. J . Bernstein and W. G. Schneider, J , Am, Chem. S o c , 8 0 , 6098 (1958). 124. M. Karplus, J . Am. Chem. S o c , 8 5 , 2870 (1963). 125. J . A. Pople, W. G. Scheider and H. J . Bernstein, High Resolution Magnetic Resonance, McGraw-Hill Book company, i n c . , New YorK, 1 9 5 9 , P. 1 3 2 . 126. P. K. W. Woo, H. W. Dion and L. F. Johnson, J , |jm, Chem. S o c , 84, 1066 (1962) . 127. L. D. H a l l , Advances i n Carbohydrate Chem., 19, i n press. 128. R. U. Lemieux and J . Howard, Can, J , Chem., 4 l , 393 (1963). 129. R. U. Lemieux and S. Levine, Can. J . Chem., 42, 1473 (1964). - 195' - 130. R. U. Lemieux and J . W. Lown, Can. J . Chem., 42, 893 (1964) . = 131. R . J . Abraham, R. Freeman, L. D. H a l l and K. A. McLauchlan, J . Chem. S o c , 2080 (1962) . 132. R. A. Edington, E. L. Hirs t and E. E. P e r c i v a l , J . Chem. S o c , 2281 (1955). 133. P. A. Levene and R. S. Tipson, J . B i o l . Chem., 9 3 , 631 (1931). = 13**. F. J . Bates and Associates, "Polarimetry, Sa c char-* met ry and the Sugars", Nat. Bur. Standard C i r c u l a r C44o (1942), P. 532. = 135. R. Criegee, L. Kraft and B. Rank, Ann., 507, 159 (1933). 136. J . W. Pratt, N. K. Richtmyer and C. S. Hudson, J . Am. Chem. S o c , 7*V, 2200 (1952). 137. J . Honeyman and C. J . G. Shaw, J . Chem. S o c , 2454 (1959). 138. S. J . Ahgyal and C. G. Macdonald, J . Chem. S o c , 686 (1952) . 139. R. J . Abraham, L. D. H a l l , L. Hough and K. A. McLauchlan, Chem. and Ind., 213 (1962)1 J . Chem. S o c , 3699 (1962). 140. R. TJ. Lemieux, J . D. Stevens and R. R. Fraser, Can. J . Chem., 40, 1955 (1962). 141. A."B. Foster, W. G. Overend, M. Stacey and L. F. Wiggins, J. Chem. S o c , 2542 (1949). 142. J . A. M i l l s , Advances i n Carbohydrate Chem., 10, 1 (1955). 143. A. T. Ness, R. M. Hann and C. S. Hudson, J . Am. Chem. Soc., 6 6 , 1901 (1944) . 144. R. M. Hann and C. S. Hudson, J . Am. Chem. S o c , 67 , 602 (1945) . == 145. W. T.'Raskins, R. M. Hann and C. S. Hudson, J. Am. Chem. S o c , 6 5 , 67 (1943) . 146. J . C. Sowden, Advances i n Carbohydrate Chem., 6 , 291 (1951). = 147. H. Ledderhose, Z. phy s i o l . Chem., 4 , 139 (1880) . 148. B. C. Bera, A. B. Poster and M. Stacey, J , Chem. S o c , 4531 (1956). 3.49. S. Akiya and T. Osawa, J . Pharm. S o c Japan, 74, 1259 (1954). — 150. P. A. Levene and P. B. LaPorge, J. B i o l . Chem., 2 1 , 345, 351 (1915). * — 151. P. A. Levene, J , B i o l , Chem., 39, 6 9 , (1919). 152. J . C. Sowden and M. T. Oftedahl, Methods i n Carbohydrate Chemistry, V o l . I, ed. by R. L. Whistler and M. L. woirrom, New York, Academic Press, 1962, p. 235. 153. A. B. Grant, New Zealand J . S c l , Techno!., 37, 509 (1956). 154. Reference 104, p. 299. 155. A. Rosenthal and H. J . Koch, unpublished r e s u l t s . 156. A. Camerman and J . Tr o t t e r , Acta Cryst., i n press. 157. M. Somogyi, J , B i o l , Chem., 16Q, 69 (1945) . 158. R. E. J . MItchel, B.Sc Thesis, University of B r i t i s h Columbia (1963). 159. Reference 79 , p. 62 . 160. R. L. Shriner, R. C. Fuson and D. Y. Curtin, Systematic I d e n t i f i c a t i o n of Organic Compounds, 4 t h Ed., Mew York, John Wiley and Sons, 1956, p. 98 . 161. D. Y. Curtin, J . A. Gourse, W. A. Richardson and K. L. Rinehart, J. Org, Chem,, 24, 93 (1959). 162. V. R. Mattox and E. C. Kendall, J . Am. Chem. S o c , 70, 882 (1948) J J , B i o l . Chem., 188, 287 ( 1 9 b D . 163. M. L. Wolfrom, J , Am, Chem, S o c , 5 3 , 2275 (1931) . 164. M. L. Wolfrom and W. M. Morgan, J , Am. Chem, S o c , 54* 3390 (1932) . — 165. N. Kornblum, W. J . Jones and G. J . Anderson, J . Am. Chem. S o c , 81 , 4113;-(1959). 166. K. E. Pf l t z n e r and J . G. Moffatt, J , Am, Chem, S o c , 8 5 , (1963). — - 197 - 167. D. H. R. Barton, B. J . Garner and R. H. Wightman, J . Chem. S o c , 1 8 5 5 ( 1 9 6 4 ) . 168. G. Henseke and G. Hanlsch, Angew. Chemle, 7 5 , 4 2 0 (1963).

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
China 4 0
Turkey 1 0
United States 1 1
City Views Downloads
Beijing 4 0
Unknown 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items